Genome Sequence and QTL Analyses Using Backcross Recombinant Inbred Lines (BILs) and BILF1 Lines Uncover Multiple Heterosis-related Loci

Heterosis is an interesting topic for both breeders and biologists due to its practical importance and scientific significance. Cultivated rice (Oryza sativa L.) consists of two subspecies, indica and japonica, and hybrid rice is the predominant form of indica rice in China. However, the molecular mechanism underlying heterosis in japonica remains unclear. The present study determined the genome sequence and conducted quantitative trait locus (QTL) analysis using backcross recombinant inbred lines (BILs) and BILF1 lines to uncover the heterosis-related loci for rice yield increase under a japonica genetic background. The BIL population was derived from an admixture variety Habataki and japonica variety Sasanishiki cross to improve the genetic diversity but maintain the genetic background close to japonica. The results showed that heterosis in F1 mainly involved grain number per panicle. The BILF1s showed an increase in grain number per panicle but a decrease in plant height compared with the BILs. Genetic analysis then identified eight QTLs for heterosis in the BILF1s; four QTLs were detected exclusively in the BILF1 population only, presenting a mode of dominance or super-dominance in the heterozygotes. An additional four loci overlapped with QTLs detected in the BIL population, and we found that Grains Height Date 7 (Ghd7) was correlated in days to heading in both BILs and BILF1s. The admixture genetic background of Habataki was also determined by subspecies-specific single nucleotide polymorphisms (SNPs). This investigation highlights the importance of high-throughput sequencing to elucidate the molecular mechanism of heterosis and provides useful germplasms for the application of heterosis in japonica rice production.


Introduction
Food security is a major global problem as the competition for arable land between food and energy crops, and population growth, continue to increase. Improving crop productivity has been the key focus of national and international efforts in breeding crops such as maize and rice. The vigor of indica rice hybrids often exhibits phenotypes that surpass their parents in terms of growth and fertility, which is also known as heterosis. In crop production, successful agronomic exploitation of yield Int. J. Mol. Sci. 2020, 21, 780 2 of 10 heterosis has been achieved in the past decades. Three major competing but non-mutually-exclusive hypotheses (dominance, overdominance, and epistasis) have been proposed to explain heterosis at the genetic level [1][2][3][4][5][6]. However, the progress in elucidating the molecular mechanism underlying crop heterosis has lagged.
In the past decades, heterosis has become a priority research target for both breeders and scientists based on its practical importance and scientific significance. Heterosis in crop breeding was first applied to hybrid maize in the 1930s, and the three-line method or cytoplasmic male sterility (CMS) system contributed to the commercialization of hybrid rice in the 1970s. Crop heterosis has been extensively applied to rice and maize production, significantly improving global yield compared with traditional inbred lines [7]. Currently, the planting area of hybrid rice accounts for 50-60% of the total rice planting area, and about 80% of indica rice is hybrid and is mainly planted in southern China. A study used recombinant inbred lines (RIL) derived from a cross between PA64s and 93-11, and RIL backcross F 1 populations were analyzed to elucidate the molecular mechanism of heterosis among indica and javanica varieties [8]. A mega sequence project for 10,0742 F 2 lines revealed the genomic architecture of heterosis for yield traits in rice using the cross combination of Oryza sativa subspecies (ssp.) indica-indica (three-line system), indica-indica (two-line system), and O. sativa ssp. indica-O. sativa ssp. japonica crosses [9]. Nevertheless, research on heterosis in japonica is limited. Almost all japonica varieties are conventional rice, and the hybrid japonica accounts for less than 3% among total japonica, which is mainly distributed in northern China and the middle range of China. The disappearance of hybrid japonica rice is due to the lack of genetic diversity among japonica cultivars.
In this study, we used a backcross RIL derived from the cross between Habataki (an admixture variety between the indica and japonica variety) and japonica variety Sasanishiki, and F 1 plants were backcrossed to Sasanishiki once before inbreeding. The backcross improved genetic diversity through the introgression of the indica pedigree while maintaining the population under the japonica genetic background. After inbreeding for 10 generations, we obtained 85 lines of the backcross recombinant inbred line (BIL) population. Then, we crossed all of the 85 lines to Sasanishiki again to generate the BILF 1 population. The present study used genome sequencing and quantitative trait locus (QTL) analysis of BILs and BILF 1 s to identify heterosis-related loci for yield increase.

Population Sequencing and Linkage Map Construction
We used a strategy of sequencing-based map construction to conduct QTL mapping for the BIL derived from the cross between Sasanishiki and Habataki ( Figure 1A). We sequenced a segregating population of Sasanishiki and Habataki BILs together with parental lines on an Illumina HiSeq2500 platform. A total of 224.02 GB of raw data were generated for all of the BILs, with approximately 6.29-fold depth for each BIL, 22.94 GB for Habataki (51.00-fold), and 23.24 GB for Sasanishiki (56.00-fold). We aligned the sequence data to the reference genome (Os-Nipponbare-Reference-IRGSP-1.0) using SOAP2 software [10,11]. A total of 1,947,668 single-nucleotide polymorphisms (SNPs) with homozygous genotypes between both parents were identified using the SOAPsnp software [12]. These SNPs were used as potential markers in the subsequent analysis. The SNP markers localized to highly repetitive regions, and those with low genotyping scores were removed to avoid ambiguity in linkage map construction. We used an effective imputation model, k-nearest neighbor algorithm to impute the missing genotypes of each RIL caused by low-coverage sequencing [13]. Finally, we used 1,591,495 high-quality polymorphic SNP markers to construct a recombinant bin map ( Figure 1B). Subsequently, a recombinant bin map was constructed, and the map contained 3652 recombinant blocks, with the average genetic length of 0.44 cM. Then, we determined the introgression rate of the Habataki pedigree using the data of bin map. The introgression rate showed a normal distribution that indicated that the population is ideal for the subsequent survey ( Figure 1C). We analyzed the correlation between the introgression rate of Habataki and the important yield-related traits. The results showed that the introgression rate was significantly negatively correlated to 1000-grain weight (TGW) ( Figure 1D). results showed that the introgression rate was significantly negatively correlated to 1000-grain weight (TGW) ( Figure 1D).

Grain Yield Heterosis is the Result of Hybrid Vigor in Grain Number Per Panicle
To obtain ideal agronomic characteristics of heterosis, the paternal line itself is usually an excellent inbred variety with superior agronomic performance. In addition, the hybrid F1 plant should surpass its paternal parent, particularly the traits related to yield components. Interestingly, the F1 plants did not exhibit greater plant height compared with the parental lines in the present study ( Figure 2A). We surveyed the yield components of the parental lines and F1 plants ( Figure 2). The panicle length of the F1 plants was significantly longer than that of Sasanishiki and Habataki ( Figure  2B). Sasanishiki exhibited a short round grain shape, whereas the F1 and Habataki exhibited a slender grain shape. However, the 1000-grain weight of Habataki was significantly lower than that of the F1 plants, whereas the 1000-grain weight of Sasanishiki was similar to that of the F1 plants. The results showed that the F1 plants exhibited an advantage in grain number per panicle and panicle number compared with Sasanishiki and Habataki. A similar setting rate was observed in F1, Sasanishiki, and Habataki. The slenderest grain shape of Habataki had the lowest 1000-grain weight in the parent line and F1 plants, whereas Sasanishiki had similar 1000-grain weight as the F1 plants. Taken together, the advantage in panicle number and grain number per panicle makes the F1 plants exhibit heterosis in grain yield per plant.

Grain Yield Heterosis is the Result of Hybrid Vigor in Grain Number Per Panicle
To obtain ideal agronomic characteristics of heterosis, the paternal line itself is usually an excellent inbred variety with superior agronomic performance. In addition, the hybrid F 1 plant should surpass its paternal parent, particularly the traits related to yield components. Interestingly, the F 1 plants did not exhibit greater plant height compared with the parental lines in the present study ( Figure 2A). We surveyed the yield components of the parental lines and F 1 plants ( Figure 2). The panicle length of the F 1 plants was significantly longer than that of Sasanishiki and Habataki ( Figure 2B). Sasanishiki exhibited a short round grain shape, whereas the F 1 and Habataki exhibited a slender grain shape. However, the 1000-grain weight of Habataki was significantly lower than that of the F 1 plants, whereas the 1000-grain weight of Sasanishiki was similar to that of the F 1 plants. The results showed that the F 1 plants exhibited an advantage in grain number per panicle and panicle number compared with Sasanishiki and Habataki. A similar setting rate was observed in F 1 , Sasanishiki, and Habataki. The slenderest grain shape of Habataki had the lowest 1000-grain weight in the parent line and F 1 plants, whereas Sasanishiki had similar 1000-grain weight as the F 1 plants. Taken together, the advantage in panicle number and grain number per panicle makes the F 1 plants exhibit heterosis in grain yield per plant.

Yield Components and Other Important Agronomic Traits of BILs and BILF1s
To assess the distribution pattern of yield-related traits between BILs and BILF1s, we surveyed the yield components of BILs and BILF1s in the paddy field of Shenyang Agricultural University in 2018. The normal distribution and transgressive segregation were observed in all of the traits surveyed in both BIL and BILF1s. These results indicate that all of these agronomic traits were controlled by multiple genes. The plant height of BILF1s mainly ranged from 105 to 120 cm, whereas that of BILs was from 120 to 135 cm. These results indicate that the heterozygous genotype of the F1 plants reduces the plant height, based on the fact that the F1 plants were shorter than Sasanishiki and Habataki ( Figure 1A). In addition, the grain number per panicle of BILs ranged from 80 to 180, whereas that of the BILF1s ranged from 130 to 260. These results suggest that the heterozygous genotype of the F1 plants increases the grain number per panicle, which also coincides with the result that the F1 plants have significantly more grains per panicle than Sasanishiki and Habataki ( Figure 2). Similar distributions of days to heading, panicle number, and 1000-grain weight were observed in both BILs and BILF1s. However, the variation range of setting rate in BILs was larger than in the BILF1s (Figure 3).

Yield Components and Other Important Agronomic Traits of BILs and BILF 1 s
To assess the distribution pattern of yield-related traits between BILs and BILF 1 s, we surveyed the yield components of BILs and BILF 1 s in the paddy field of Shenyang Agricultural University in 2018. The normal distribution and transgressive segregation were observed in all of the traits surveyed in both BIL and BILF 1 s. These results indicate that all of these agronomic traits were controlled by multiple genes. The plant height of BILF 1 s mainly ranged from 105 to 120 cm, whereas that of BILs was from 120 to 135 cm. These results indicate that the heterozygous genotype of the F 1 plants reduces the plant height, based on the fact that the F 1 plants were shorter than Sasanishiki and Habataki ( Figure 1A). In addition, the grain number per panicle of BILs ranged from 80 to 180, whereas that of the BILF 1 s ranged from 130 to 260. These results suggest that the heterozygous genotype of the F 1 plants increases the grain number per panicle, which also coincides with the result that the F 1 plants have significantly more grains per panicle than Sasanishiki and Habataki ( Figure 2). Similar distributions of days to heading, panicle number, and 1000-grain weight were observed in both BILs and BILF 1 s. However, the variation range of setting rate in BILs was larger than in the BILF 1 s (Figure 3).

QTL Detection and Analysis Using the BILs and BILF1 Population
To elucidate the genetic mechanism underlying heterosis of yield traits, we primarily focused on the days to heading (DTH), plant height (PH), panicle number (PN), grain number per panicle (GNPP), setting rate (SR), and 1000-grain weight (TGW). A molecular linkage map with 3652 bins was constructed based on sequence variations in Sasanishiki and Habataki and those among the BILs using Highmaps software. In the BIL population, 10 QTLs for all of the traits were mapped independently to rice chromosomes 1, 4, 5, 7, 9, 10, and 12 ( Figure 4). To detect QTLs for heterosis that was associated with the effects of the heterozygous on the genetic background of Sasanishiki, we used the BILF1 to conduct QTL analysis. Eight loci for the respective phenotypes were detected on chromosomes 1, 3, 6, 7, 9, and 10. Among the eight QTLs, four of these loci overlapped with QTLs detected in the BIL population, and four QTLs were detected independently in the BILF1 population. These results presented a mode of dominance or super-dominance in the heterozygote. Among the QTLs detected in both BILs and BILF1S, the cluster at the short arm of chromosome 10 corresponded to the days to heading in BILF1s and the grain number per panicle in both BILs and BILF1s. In addition, a cluster at chromosome 7 was related to days to heading in both BILs and BILF1s. As heterosis was mainly observed in the grain number per panicle, we subsequently summarized the grain number per panicle of each genotype for the four QTLs detected in the BIL and BILF1s ( Figure  5). The results showed that in the Gn1 and Ghd7 loci, the heterozygous genotype and the homozygous Habataki genotype plants had significant advantage in grain number per panicle compared with the homozygous Sasanishiki genotype plants. In the GNNP3 locus, the homozygous Sasanishiki and homozygous Habataki genotype plants had similar grain number per panicle but significantly fewer than the heterozygous genotype plants. In the Gn3 locus, the heterozygous genotype plants showed a significant increase of grain number per panicle compared with homozygous Habataki genotype plants, whereas the homozygous Habataki genotype plants had a significantly higher grain number per panicle than the homozygous Sasanishiki genotype plants.

QTL Detection and Analysis Using the BILs and BILF 1 Population
To elucidate the genetic mechanism underlying heterosis of yield traits, we primarily focused on the days to heading (DTH), plant height (PH), panicle number (PN), grain number per panicle (GNPP), setting rate (SR), and 1000-grain weight (TGW). A molecular linkage map with 3652 bins was constructed based on sequence variations in Sasanishiki and Habataki and those among the BILs using Highmaps software. In the BIL population, 10 QTLs for all of the traits were mapped independently to rice chromosomes 1, 4, 5, 7, 9, 10, and 12 ( Figure 4). To detect QTLs for heterosis that was associated with the effects of the heterozygous on the genetic background of Sasanishiki, we used the BILF 1 to conduct QTL analysis. Eight loci for the respective phenotypes were detected on chromosomes 1, 3, 6, 7, 9, and 10. Among the eight QTLs, four of these loci overlapped with QTLs detected in the BIL population, and four QTLs were detected independently in the BILF 1 population. These results presented a mode of dominance or super-dominance in the heterozygote. Among the QTLs detected in both BILs and BILF 1S , the cluster at the short arm of chromosome 10 corresponded to the days to heading in BILF 1 s and the grain number per panicle in both BILs and BILF 1 s. In addition, a cluster at chromosome 7 was related to days to heading in both BILs and BILF 1 s. As heterosis was mainly observed in the grain number per panicle, we subsequently summarized the grain number per panicle of each genotype for the four QTLs detected in the BIL and BILF 1s ( Figure 5). The results showed that in the Gn1 and Ghd7 loci, the heterozygous genotype and the homozygous Habataki genotype plants had significant advantage in grain number per panicle compared with the homozygous Sasanishiki genotype plants.
In the GNNP3 locus, the homozygous Sasanishiki and homozygous Habataki genotype plants had similar grain number per panicle but significantly fewer than the heterozygous genotype plants.
In the Gn3 locus, the heterozygous genotype plants showed a significant increase of grain number per panicle compared with homozygous Habataki genotype plants, whereas the homozygous Habataki genotype plants had a significantly higher grain number per panicle than the homozygous Sasanishiki genotype plants.  The number inside the column represents the number of plants for each genotype. All data are given as mean ± s.e.m.,different letters indicate the significant differences at the 5% level.

The Indica/Japonica Pedigree Analysis of Sasanishiki and Habataki
Breeding history indicates that Habataki is an admixture line between indica and japonica. Thus we conducted an analysis of the indica/japonica pedigree of Sasanishiki and Habataki. The indica/japonica pedigree was defined using the subspecies-specific SNPs. The subspecies-specific SNPs were those of the same type in all japonica, but not in indica, which is based on the divergence of the 517 rice landraces [14]. In total, 100,529 subspecies-specific SNPs were selected. We matched the 1,947,668 SNPs between Sasanishiki and Habataki to 100,529 subspecies-specific SNPs, and 81,690 SNPs were merged. The 81,690 SNPs were then used to analyze the indica/japonica pedigree of Sasanishiki and Habataki. The results showed that there are 621 japonica-type SNPs in the genome of Habataki, which indicated that 0.76% of japonica genomic introgression involved Habataki. Meanwhile, there were only 81 indica-type SNPs in Sasanishiki,   The number inside the column represents the number of plants for each genotype. All data are given as mean ± s.e.m.,different letters indicate the significant differences at the 5% level.

The Indica/Japonica Pedigree Analysis of Sasanishiki and Habataki
Breeding history indicates that Habataki is an admixture line between indica and japonica. Thus we conducted an analysis of the indica/japonica pedigree of Sasanishiki and Habataki. The indica/japonica pedigree was defined using the subspecies-specific SNPs. The subspecies-specific SNPs were those of the same type in all japonica, but not in indica, which is based on the divergence of the 517 rice landraces [14]. In total, 100,529 subspecies-specific SNPs were selected. We matched the 1,947,668 SNPs between Sasanishiki and Habataki to 100,529 subspecies-specific SNPs, and 81,690 SNPs were merged. The 81,690 SNPs were then used to analyze the indica/japonica pedigree of Sasanishiki and Habataki. The results showed that there are 621 japonica-type SNPs in the genome of Habataki, which indicated that 0.76% of japonica genomic introgression involved Habataki. Meanwhile, there were only 81 indica-type SNPs in Sasanishiki, Figure 5. Grain number per panicle of each genotype for the four QTLs detected in BIL and BILF 1s . The number inside the column represents the number of plants for each genotype. All data are given as mean ± s.e.m.,different letters indicate the significant differences at the 5% level.

The Indica/Japonica Pedigree Analysis of Sasanishiki and Habataki
Breeding history indicates that Habataki is an admixture line between indica and japonica. Thus we conducted an analysis of the indica/japonica pedigree of Sasanishiki and Habataki. The indica/japonica pedigree was defined using the subspecies-specific SNPs. The subspecies-specific SNPs were those of the same type in all japonica, but not in indica, which is based on the divergence of the 517 rice landraces [14]. In total, 100,529 subspecies-specific SNPs were selected. We matched the 1,947,668 SNPs between Sasanishiki and Habataki to 100,529 subspecies-specific SNPs, and 81,690 SNPs were merged. The 81,690 SNPs were then used to analyze the indica/japonica pedigree of Sasanishiki and Habataki. The results showed that there are 621 japonica-type SNPs in the genome of Habataki, which indicated that 0.76% of japonica genomic introgression involved Habataki. Meanwhile, there were only 81 indica-type SNPs in Sasanishiki, indicating that only 0.01% indica pedigree introgression occurred in the Sasanishiki genome. The distribution of subspecies-specific SNPs and indica/japonica genomic introgression is shown in Figure 6. Agronomic-traits-related QTLs were located out of the indica/japonica genomic introgression, which confirmed that the heterosis QTL originated from the difference between indica and japonica.
indicating that only 0.01% indica pedigree introgression occurred in the Sasanishiki genome. The distribution of subspecies-specific SNPs and indica/japonica genomic introgression is shown in Figure 6. Agronomic-traits-related QTLs were located out of the indica/japonica genomic introgression, which confirmed that the heterosis QTL originated from the difference between indica and japonica.

Discussion
Since its generation in 1973, hybrid rice has predominated indica rice production in China. Numbers of elite combinations have been developed and released as commercial varieties, with yields roughly 20% higher than their inbred counterparts [15]. Recent molecular research has investigated the number of hybrid combinations to construct a model system for studying the molecular mechanism of heterosis for three-and two-line hybrids. A study using the yield components data and an ultra-high-density SNP bin map of an immortalized F2 population derived

Discussion
Since its generation in 1973, hybrid rice has predominated indica rice production in China. Numbers of elite combinations have been developed and released as commercial varieties, with yields roughly 20% higher than their inbred counterparts [15]. Recent molecular research has investigated the number of hybrid combinations to construct a model system for studying the molecular mechanism of heterosis for three-and two-line hybrids. A study using the yield components data and an ultra-high-density SNP bin map of an immortalized F 2 population derived from the cross between Zhenshan97 and Minghui63 demonstrated that the relative contributions of the genetic components vary with traits.
The results indicate that overdominance/pseudo-overdominance are most important to the heterosis of grain number per panicle, 1000-grain weight, and grain yield per plant. In heterosis of panicle number and 1000-grain weight, the dominance × dominance interaction is important. Among these yield-related traits, single-locus dominance has relatively small contributions [16]. An integrated analysis using the RIL population and RILBCF 1 population derived from the cross between PA64S (which has a mix genetic background of indica and javanica) and 93-11 showed that heterosis was mainly detected in grain number per panicle and panicle number [8]. Huang et al. sequenced 10,074 F 2 lines derived from 17 representative hybrid combinations, and they found that a small number of genomic loci from female parents explain a large proportion of the yield advantage of hybrids over their male parents [9]. Taken together, these studies improve our understanding of heterosis. However, research on heterosis between japonica and japonica crosses is limited. The genetic diversity of japonica was not as high as that of indica, making heterosis among japonica inconspicuous compared with that among indica. Moreover, several factors have limited the utilization of heterosis between indica and japonica, which includes sterility. Thus, we used a BIL population derived from an indica and japonica cross to improve the genetic diversity but maintain the genetic background close to japonica. The results of the present study showed that heterosis between Sasanishiki and Habataki mainly came from a complicated quantitative and components-specific phenotype. Here, we clearly demonstrate yield heterosis, mainly by the outperformance of grain number per panicle and panicle number. Yield components survey also showed that the grain number per panicle of BILF 1 population was significantly higher than that of the BIL population. Our previous study demonstrated that the introgression of indica pedigree in the japonica genome contributed to the increase of rice production in northern China [17], and thus the present study confirmed that the indica pedigree could increase the grain number per panicle in japonica.
Using high-throughput sequencing, we conducted QTL mapping of the BIL and BILF 1 populations, and the number of QTLs that are responsible for yield and yield-related heterosis was determined. A total of 10 QTLs for all of the traits were mapped independently in BIL population (Figure 4), and eight loci for the respective phenotypes were detected. Among the eight QTLs, four of these loci overlapped with QTLs detected in the BIL population, and the remaining four QTLs were detected exclusively in the BILF 1 population, indicating a mode of dominance or super-dominance in the heterozygote. At the gene level, Gao et al. demonstrated that a heading-time-regulated gene Day to Heading 8 (DTH8) is a candidate locus for yield heterosis in Liang-you-pei 9 (LYP9) [18], and Li et al. and Huang et al. confirmed that DTH8 corresponds to yield heterosis [8,9]. These results suggest that the heading time gene strongly participates in yield heterosis in hybrid rice. The present study detected that the heading-time gene Ghd7 is also a candidate locus for both yield heterosis and heading time regulation. A study using the BILs derived from the cross between Habataki and Koshihikari identified five QTLs (Gn1-Gn5) that were related to grain number per panicle on chromosome 1, 4, 10, and 12 [19]. The present study confirmed that Gn1 and Gn3 correspond to grain number per panicle. Moreover, Gn3, on the short arm of chromosome 10, also corresponded to grain number per panicle in BILF 1 s. Thus, this QTL may be further assessed in terms of heterosis.

Plant Materials
The parental line Sasanishiki is typical japonica varieties, and the parental line Habataki is an admixture variety with both indica and japonica pedigree. Sasanishiki was crossed to Habataki, and the F 1 plant was crossed to Sasanishiki, and then inbred over 10 generations by single-seed descent to generate a population containing 85 BILs. The BILs were then backcrossed to the maternal parent Sasanishiki to obtain 85 BILF 1 s. All materials are maintained at the Rice Research Institute of Shenyang Agricultural University (Shenyang, China). All plants were planted in random block design under standard agricultural management practice in the paddy field of Shenyang Agricultural University (N41 • , E123 • ). All lines were planted with three biological replicates A total of 300 plants in a 16 m 2 plot were characterized for each line. Plant height, panicle number, grain number per panicle, setting rate, and 1000-grain weight were surveyed in the field in 2018. Both BILs and BILF 1 s were used in QTL analysis.

DNA Extraction and QTL Analysis
Young leaves of each line were collected two weeks after transplant. The CTAB method was used in extracting the high-quality genomic DNA. The sequencing libraries were constructed on an Illumina HiSeq2500 platform (Illumina, Inc.; San Diego, CA, USA) according to the manufacturer's instructions. We aligned the sequence data to the reference genome (Nipponbare, http://rapdb.dna. affrc.go.jp/download/irgsp1.html/) using SOAP2 software [11]. To construct the genetic linkage map for QTL analysis, we combined the cosegregating SNP/InDel into bins via HighMap software [20]. A map containing 3652 bins and 1592.12 cM in length was constructed with an average of 304 bins on each chromosome. Twelve linkage groups corresponded to the 12 rice chromosomes. We observed the full collinearity between the genetic map and the rice genome, and the minimum value of the Spearman coefficient for chromosomes was 0.962 (Chr. 7). The HighMap software was used to construct a linkage map. The software constructs high-quality linkage map according to the maximum likelihood estimation method. We used the R/qtl (version: 1.44-9) software to conduct QTL analysis via a composite interval mapping (CIM) model. The significance thresholds were determined by 1000 permutations. The percentage of phenotypic variance calculation explained by each QTL was obtained according to the population variance within the mapping population. The details of the QTL analysis were described in our previous studies [21].