Mapping Quantitative Trait Loci (QTLs) for Hundred-Pod and Hundred-Seed Weight under Seven Environments in a Recombinant Inbred Line Population of Cultivated Peanut (Arachis hypogaea L.)

The cultivated peanut (Arachis hypogaea L.) is a significant oil and cash crop globally. Hundred-pod and -seed weight are important components for peanut yield. To unravel the genetic basis of hundred-pod weight (HPW) and hundred-seed weight (HSW), in the current study, a recombinant inbred line (RIL) population with 188 individuals was developed from a cross between JH5 (JH5, large pod and seed weight) and M130 (small pod and seed weight), and was utilized to identify QTLs for HPW and HSW. An integrated genetic linkage map was constructed by using SSR, AhTE, SRAP, TRAP and SNP markers. This map consisted of 3130 genetic markers, which were assigned to 20 chromosomes, and covered 1998.95 cM with an average distance 0.64 cM. On this basis, 31 QTLs for HPW and HSW were located on seven chromosomes, with each QTL accounting for 3.7–10.8% of phenotypic variance explained (PVE). Among these, seven QTLs were detected under multiple environments, and two major QTLs were found on B04 and B08. Notably, a QTL hotspot on chromosome A08 contained seven QTLs over a 2.74 cM genetic interval with an 0.36 Mb physical map, including 18 candidate genes. Of these, Arahy.D52S1Z, Arahy.IBM9RL, Arahy.W18Y25, Arahy.CPLC2W and Arahy.14EF4H might play a role in modulating peanut pod and seed weight. These findings could facilitate further research into the genetic mechanisms influencing pod and seed weight in cultivated peanut.


Introduction
The cultivated peanut (Arachis hypogaea L.), an allotetraploid (2n = 4x = 40) crop, is an important oil crop worldwide [1].Global annual peanut production is on a rapid upward trend, from 2015 (45 million tons) to 2021 (53.9 million tons) (FAO, 2021), but still cannot satisfy the demand of the growing global population.Increasing peanut yield per unit area remains a tremendous challenge for peanut breeders.Various agronomic characteristics, including total branch numbers (TBNs), lateral branch angle (LBA) and the size of the pod and seed, affect the yield of peanuts [2][3][4][5].HPW and HSW, which are mainly determined by pod and seed weight and size, are vital components.They are typical quantitative traits, but their underlying genetic basis is yet to be thoroughly researched [6].By creating a HDGM and identifying QTLs for HPW and HSW, along with mining molecular markers closely associated with yield traits, a theoretical groundwork can be laid for further revealing the genetic basis of yield traits, enhancing peanut production.
Quantitative trait loci for yield-related characteristics were identified using segregated populations in peanuts, such as PL (pod length), PW (pod width), SL (seed length), SW (seed width), HPW (hundred-pod weight) and HSW (hundred-seed weight) [24,[29][30][31][32]. Nevertheless, as the two key investigation traits of pod and seed weight and size, the majority of QTLs for the two traits showed a small effect, with a phenotypic variation explained (PVE) result of less than 10%.Up to now, Luo et al. [33] have employed a RIL population (Yuanza 9102 × Xuzhou 68-4), and identified three major effective QTLs affecting HPW across four seasons.Wang et al. [23] established a RIL (ZH16 × sd-H1) and obtained two QTLs for HPW (5.86-14.46% of PVE) and six QTLs for HSW (5.17-17.95% of PVE) under three environments.Mondal et al. [34] utilized a RIL population (VG 9514 × TAG 24), leading to the identification of nine QTLs for HSW spread over six environments, which accounted for 6.71-23.88% of PVE.Kunta et al. [24] developed a RIL (Hanoch × Harariused) and discovered 30 QTLs across two seasons, including eight QTLs for 50-pod weight and -seed weight.Of these, three exhibited main effects, explaining 6.2 to 13.9% of PVE.Similarly, Gangurde et al. [35] also used a RIL population (Chico × ICGV02251) and detected seven QTLs associated with HSW in three years, responsible for 6.69-21.29% of PVE.Guo et al. [36], using a RIL derived from Zhonghua 5 and ICGV 86699, identified 15 QTLs for HSW across six settings, explaining a phenotypic variation comprising 4.08-17.89%.Within these QTLs, only three displayed main effects.Nevertheless, most QTLs cannot be detected repeatedly in multiple environments and the number of QTLs showing stable expression is still relatively low.Stable QTLs are those that have been consistently detected across multiple years in multiple environments.Obviously, there is still a lack of main-effect QTLs for stable expression in multiple environments.
Stable expression of QTLs in multiple environments is important for revealing the genetic mechanisms of crop growth and development.In our current research, to further elucidate the candidate regions of the genomic impact of pod and seed weight in peanuts, we developed a RIL population of 188 families, achieved by crossing two cultivated peanut species."Jihua5 (JH5)", as the female receptor, was a large-seed and erect plant type, and "M130,"as the male donor, was a small-seed with spreading plant type.HPW and HSW were significantly different between the two parents, and they presented plentiful variations in RIL generation; thus, they were suitable for QTL localization.Here, we gathered genotype data of SSR, AhTE, SRAP, TRAP and SNP markers to construct a novel HDGM.To test the practicability of the map, QTLs for the HPW and HSW were mapped across seven environments over four years.Interestingly, a QTL hotspot was discovered on chromosome A08, which holds potential significance for the future breeding of peanut pod and seed traits.

Plant Materials and Multiple Environment Trials
A RIL population was established through the F 8:11 generation from a cross between "JH5" and "M130" using the single seed descent (SSD) method, a high-density genetic linkage map was constructed, and QTL analysis was performed for both HPW and HSW.JH5, as a female parent, was a peanut cultivar with large pods and seeds.M130, as a male parent, was a peanut germplasm with small pods and seeds (Figure 1).The 188 RILs and their parents were planted under seven environments over four years, including the Qingyuan experimental field (QY, N38 • 40 and E115 • 30 ) in the years 2017, 2018 and 2020, Daming (DM, N35 • 57 and E115 • 09 ) in the years 2017 and 2018, Qian'an (QA, N39 • 99 and E118 • 70 ) in 2018, and Xinle (XL, N38 • 15 and E114 • 30 ) in 2019, which were referred to as 17QY, 17DM, 18QY, 18DM, 18QA, 19XL and 20QY, respectively.We applied a randomized block design to the 188 lines with two replications, and crop field management followed local requirements.Each plot, with 10 plants, was grown in one row, the row length, row spacing and planted spacing of each one was 1.8 m, 0.5 m and 0.2 m, respectively.The parental lines were planted after every 20 rows as controls.Planting of seeds took place in May and harvest occurred in September for each experiment.

Plant Materials and Multiple Environment Trials
A RIL population was established through the F8:11 generation from a cross between "JH5" and "M130" using the single seed descent (SSD) method, a high-density genetic linkage map was constructed, and QTL analysis was performed for both HPW and HSW.JH5, as a female parent, was a peanut cultivar with large pods and seeds.M130, as a male parent, was a peanut germplasm with small pods and seeds (Figure 1).The 188 RILs and their parents were planted under seven environments over four years, including the Qingyuan experimental field (QY, N38°40′ and E115°30′) in the years 2017, 2018 and 2020, Daming (DM, N35°57′and E115°09ʹ) in the years 2017 and 2018, Qian'an (QA, N39°99′ and E118°70′) in 2018, and Xinle (XL, N38°15′ and E114°30′) in 2019, which were referred to as 17QY, 17DM, 18QY, 18DM, 18QA, 19XL and 20QY, respectively.We applied a randomized block design to the 188 lines with two replications, and crop field management followed local requirements.Each plot, with 10 plants, was grown in one row, the row length, row spacing and planted spacing of each one was 1.8 m, 0.5 m and 0.2 m, respectively.The parental lines were planted after every 20 rows as controls.Planting of seeds took place in May and harvest occurred in September for each experiment.

Traits Measurement and Statistical Analysis
Eight typical plants from each plot were harvested and picked ripe and plum-pod at the mature stage.HPW and HSW were evaluated utilizing an electronic balance with three replicates for accurate measurements.Data analysis was conducted using Prism 8.0 (GraphPad software), which assisted in analyzing descriptive statistics and variance components.This software was also vital in deducing the Pearson's correlation coefficient amidst HPW and HSW.To validate whether the data for the two traits conformed to a normal distribution, we employed the Shapiro-Wilk test (w-test) for the normality evaluation of the phenotypic data.The broad-sense heritability (hB 2 ) for HPW and HSW across seven environments was quantified via: hB 2 = σg 2 /(σg 2 +σge 2 /n+σe 2 /nr), where σg 2 , σge 2 and σe 2

Traits Measurement and Statistical Analysis
Eight typical plants from each plot were harvested and picked ripe and plum-pod at the mature stage.HPW and HSW were evaluated utilizing an electronic balance with three replicates for accurate measurements.Data analysis was conducted using Prism 8.0 (GraphPad software), which assisted in analyzing descriptive statistics and variance components.This software was also vital in deducing the Pearson's correlation coefficient amidst HPW and HSW.To validate whether the data for the two traits conformed to a normal distribution, we employed the Shapiro-Wilk test (w-test) for the normality evaluation of the phenotypic data.The broad-sense heritability (h B 2 ) for HPW and HSW across seven environments was quantified via: h B 2 = σ g 2 /(σ g 2 +σ ge 2 /n+σ e 2 /nr), where σ g 2 , σ ge 2 and σ e 2 symbolize the genetic variance component, genotype-environment interaction (G × E) variance component, and the random error variance component, respectively.Herein, 'n' signifies the number of environments and 'r' denotes the number of replications encompassed in each field experiment.

Marker Polymorphism and Analysis
Total genomic DNA was extracted from fresh leaves of RILs and two parents following the method of Wang et al. [23].A sum of 8091 markers was obtained to screen the polymorphism of the two parents.Among these: 2808 polymorphic SNP markers from our previous research [25], 3964 pairs of SSR and 926 transposon element markers (AhTE) (https://legacy.peanutbase.org/,accessed on 15 August 2019), 238 pairs of SRAP primers [37] and 155 pairs of TRAP primers [38].Primers were synthesized by Genewiz (Suzhou, China).The polymerase chain reaction (PCR) system of SSR and AhTE was conducted in a 10 µL mixture, including 5 µL of 2 × Es Taq Master Mix (Cwbio, Taizhou, China), 1 Ml of template DNA (10 ng/Ml), 0.5 Ml of forward and reverse primer (10 Mm/Ml), and 2 Ml double-distilled water.The PCR procedure involved the following steps: 95 • C for 5 min, then 30 cycles of 94 • C/40 s, 55 • C/40 s and 72 • C/60 s, final extension at 72 • C for 10 min and a cool-down process at 4 • C. In addition, SRAP and TRAP PCR amplification procedures were performed as described in Li and Quiro [37] and Hu and Vick [38].The PCR products were investigated using 8% non-denaturing polyacrylamide gels.Silver staining was performed as described by Yang et al. [39].Subsequently, the polymorphic markers were deployed to screen the RIL population.

Construction of Integrated Genetic Linkage Map
Combining SSR, AhTE, SRAP, TRAP in our current study and the previously reported SNP marker [25], an integrated genetic linkage map was constructed using JoinMap ® 4 [40] with a logarithm of odds (LOD) threshold of 3.0 and a maximal distance of 50 cM.The identification of segregation distortion loci was achieved by using the chi-square (χ 2 ) test, and the construction of the genetic map excluded any molecular markers deviating from the expected 1:1 ratio.Genetic map distances were calculated by the Kosambi function [41], with a recombination frequency of 0.45.The genetic linkage groups were graphically presented using Mapchart 2.32 [42].

QTL Identification and Candidate Genes Prediction for QTL Hotspot
QTL IciMapping V4.2 [43] (statistical model: ICIM-ADD) was employed to identify QTLs for HPW and HSW.For each trait, a walk step of 0.5 cM and LOD threshold were estimated by permutation test 1000 times to determine a significant QTL.The QTL nomenclature was adopted according to Tanksley and McCouch [44].A major QTL has more than 10% phenotypic variation explained (PVE) [34].QTLs in the same location or overlapping region on the same chromosome are defined as a QTL hotspot.Subsequently, the candidate genes of the QTL hotspot were found according to the physical position on the reference genome of flanking markers.Then, candidate genes were analyzed for GO and KEGG enrichment.

Phenotypic Analysis
In the two parents, "JH5" indicated greater HPW and HSW than "M130" in all seven environments, and the RIL population exhibited adequate variation types (Table 1).The distribution of HPW and HSW's median and dispersion fluctuates marginally for each environment, showed a right-skewed pattern (Figures 2 and 3).The phenotypic data of HPW and HSW were continuously distributed in the RIL population and confirmed to be normally distributed by the Shapiro-Wilk (w) test (Table 1 and Figure 2).The HPW and HSW of female parent JH5 varied from 213.2 to 291.31 g and 96.12 to 114.32 g, while the HPW and HSW of male parent M130 varied from 155.79 to 177.85 g and 63.88 to 76.2 g in the seven environments, respectively (Table 1).The HPW and HSW were significantly positively correlated in all seven environments, with a correlation coefficient range from 0.844 to 0.962 (Table 2).There were high phenotypic variations of HPW and HSW, with ranges of 71.04-239.22  , respectively.The broad-sense heritability of HPW and HSW were estimated to be 0.64 and 0.52.Analysis of variance (ANOVA) showed that genotype, environmental and genotype-by-environment interaction had a significant effect on HPW and HSW (Table 3).

Integrated Genetic Map Construction and Marker Distribution
A total of 377 SSRs (9.51%), 131 AhTEs (14.15%), 90 SRAP primer pairs (37.81%) and 42 TRAP primer pairs (27.09%) had clear bands and excellent polymorphism between JH5 and M130.These polymorphism primers were used to obtain genotype data from the RIL population.In addition, 2808 SNP genotypic data from our previous study were also deployed to create an integrated high-density genetic linkage map (IHDGM) in the present study.Finally, an IHDGM with 3130 loci, covering 1998.92cM with an average distance of 0.64 cM, was constructed on 20 chromosomes, including 2796 SNPs, 229 SSRs, 30 AhTEs, 56 SRAPs and 19 TRAPs.Of these, the "A" subgroup contained 1594 loci spanning 1038.87 cM with an average distance 0.68 cM, and the "B" subgroup contained 1536 loci spanning 960.05 cM with an average distance 0.63 cM.The length of a single linkage group ranged from 50.2 to 192 cM, and the maximum gap between markers was 20.58 cM (Tables 4 and S1, Figure 4).

QTL Hotspot and Candidate Genes on A08
Based on multi-environment QTL co-localization analysis, high LOD intervals for HPW and HSW were detected in several conditions (Figure 6A).A total of 10 QTLs (HPW for 17QY, 17DM, 18DM, 18QA, 19XL and 20QY, HSW for 18DM, 18QA, 19XL and 20QY) associated with peanut traits of HPW and HSW on chromosome A08 were mapped using the flanking markers AhTE0658 and TC22C01 (Figure 6B), covering a genetic distance of 2.75 cM.The physical locations of markers AhTE0658 and TC22C01 are 35,963,966 bp and 36,328,872 bp, respectively, spanning a physical interval of 0.36 Mb on chromosome A08.

Discussion
The completion of whole genome sequencing for the tetraploid peanut and whole genome resequencing of several cultivated varieties have minimized the likelihood of marker position discrepancies, consequently enhancing the precision of QTL/gene mapping [20,45].SNP, as a third-generation molecular marker, compared with the previous two generations, showed a more abundant polymorphism in peanut germplasm resources.Constructing a high-density genetic map through developing molecular markers led to improved efficiency and accuracy of QTL mapping of interest traits.So far, many studies have identified QTLs using genetic linkage maps of different molecular markers in the The QTL hotspot interval on A08 was mapped on 6.67-9.42cM in this map.This interval was mapped at 35,963,966-36,328,872 bp in chromosome A08 by the flanking markers AhTE0658 and TC22C01 (Figure 6C).The 0.36 Mb interval contained 18 putative genes (https://legacy.peanutbase.org/gb2/gbrowse/arahy.Tifrunner.gnm2/,accessed on 15 August 2023).Of these, annotation information of 17 genes had been described, and one gene was described as an unknown protein (Table 6).Arahy.W18Y25 and Arahy.CPLC2W encode a PPR superfamily and a PPR-like superfamily, respectively.Arahy.IBM9RL, Arahy.14EF4H and Arahy.D52S1Z encode FERTILIZATION-INDEPENDENT ENDOSPERM-like (FIE), sugar transporter 11 and 2-oxoglutarate/Fe(II)-dependent dioxygenase-like, respectively.Twelve of eighteen genes were assigned at least one GO term.These 18 genes are divided into three GO categories, cellular components with 3 genes, molecular functions with 11 genes and biological processes with 9 genes.Enrichment analysis indicated that four candidate genes were enriched, including translational initiation, cell redox homeostasis, extrinsic component of membrane, transferase activity and transferring glycosyl groups (Table S2).Through KEGG enrichment analyses, four genes were found to be involved with fatty acid elongation, diterpenoid biosynthesis, photosynthesis and fatty acid metabolism biosynthesis (Table S3).

Discussion
The completion of whole genome sequencing for the tetraploid peanut and whole genome resequencing of several cultivated varieties have minimized the likelihood of marker position discrepancies, consequently enhancing the precision of QTL/gene mapping [20,45].SNP, as a third-generation molecular marker, compared with the previous two generations, showed a more abundant polymorphism in peanut germplasm resources.Constructing a high-density genetic map through developing molecular markers led to improved efficiency and accuracy of QTL mapping of interest traits.So far, many studies have identified QTLs using genetic linkage maps of different molecular markers in the peanut [24,37,[46][47][48].The density of the genetic map increased from hundreds to thousands.However, the most high density genetic maps only contained 1 to 2 molecular marker types.For example, our previous study constructed a HDGM with 2808 SNPs [25], and another HDGM with 2996 SNPs and 330 SSRs [27].Hu et al. [26] constructed a HDGM with 68 SSRs and 2266 SNPs.To further improve the accuracy and consistency of QTL detection, we constructed an integrated high-density genetic map using five types of molecular markers in this study.This map contained 3130 loci using 2796 SNPs, 229 SSRs, 30 AhTEs, 56 SRAPs and 19 TRAPs, and is currently the most comprehensive and integrated record of marker data available.Despite our efforts, we were not able to confirm the localization markers consistent with prior published results.This could be attributed to different mapping populations or the absence of adequate map density.As a result, we are compelled to consider the necessity of further refining our integrated map.
Pod and seed weight are crucial indicators of yield and have been extensively studied in various crops [49][50][51].However, the genetic mechanism underlying these traits in peanut seeds remains unclear and requires further investigation.Up to the present date, several QTLs related to pod and seed traits have been discovered on different chromosomes in peanuts [52,53].In recent years, QTLs for pod and seed weight identified on A05, A07 and B06 have been repeatedly reported [34,54,55].Similarly, in our study, QTLs for HPW and HSW were identified on B06, except for A05 and A07.Notably, qHPWA08.3,qHPWB08.1,qHPWB08.2 and qHSWA08.6 were identified in more than three environments, suggesting the stability of their genetic effects across different conditions.In addition, two QTL clusters for HPW were found on A08.It is worth noting that qHPWA08.1,qHPWA08.2,qHPWA08.3,qHPWA08.4,qHSWA08.1,qHSWA08.2 and qHSWA08.3 were located in the 0.36 Mb genome interval, but each QTL showed a micro effect and less than 10% PVE.QTL clusters could be considered multifactorial linkages, in accordance with the polygene hypothesis, and when consistently detected across multiple environments, they show a stronger association with traits [56].Consequently, QTL clusters on A08 in the current study, which were detected repeatedly in multiple environments, were considered to be strongly correlated with pod and seed weight.Our findings offer valuable insights into the genetic basis of pod and seed weight traits in the peanut, and the suggestion that the QTL cluster related HPW and HSW on A08 enhances our confidence in the accuracy of the QTLs identified on other chromosomes.
In this study, we identified 18 potential genes situated on A08, covering a physical interval of 0.36 Mb.Of these, there were five candidate genes that might regulate seed development in the plant.For instance, 2-oxoglutarate/Fe(II)-dependent dioxygenase-like possesses intrinsic functions in DNA repair, epigenetics, and post-translational modification, as well as the activation and catabolism of plant growth regulators.Furthermore, it orchestrates the production and catabolism of numerous plant hormones, including gibberellins (GAs), ethylene, auxin (IAA) and salicylic acid (SA) [57].The FIE protein serves as a vital structural component of expected PRC2 complexes, and plays a critical role in various plant growth stages, such as seed development, the transition from the vegetative phase, and the response to vernalization [58].The PPR family of proteins is an important family of genes involved in a multitude of plant growth and development processes.PPR proteins, which bind to RNA, participate in various post-transcriptional regulatory processes, and play pivotal roles in the development of plant leaves, seed development and response to stress [59].Sugar transporter proteins play a crucial role in the maturation of cereal crops.They provide possible genetic pathways for enhancing seed filling and productivity, particularly in maize and other grain crops [60,61].In this study, genes Arahy.W18Y25 and Arahy.CPLC2W encoded the PPR and PPR-like proteins superfamilies, respectively.Arahy.IBM9RL, Arahy.14EF4H and Arahy.D52S1Z encoded FIE, sugar transporter 11 and 2-oxoglutarate/Fe(II)-dependent dioxygenase-like, respectively.Therefore, we suggested that these candidate genes may participate in regulating seed development, such as seed size and weight, and speculated that the QTL hotspot (35,963,328,872 bp) on A08 was also a vital genome region for pod and seed weight.

Conclusions
In the present study, a RIL population was constructed using female parent JH5 and male parent M130.A HDGM was constructed including a total of 3130 marker loci and spanning a 1998.92cM genetic distance.In total, 31 QTLs for HPW and HSW were detected on A03, A04, A08, B04, B05, B06 and B08.A QTL hotspot was identified on A08, which was across a 0.36 Mb physical interval and included 18 candidate genes.This work will provide favorable information for researchers to breed high-yield cultivars and an analysis of the genetic mechanisms for pod and seed weight in the peanut.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/genes14091792/s1,Table S1-Marker position information of the integrated high-density genetic map.Table S2-GO annotation of candidate regions.Table S3-KEGG annotation of candidate regions.

Figure 1 .
Figure 1.Phenotypic features of pod and seed weight of both JH5 and M130.(A) HPW of JH5.(B) HPW of M130.(C) t test for HPW of JH5 and M130.(D) HSW of JH5.(E) HSW of M130.(F) t test for HSW of JH5 and M130.Scale bar was 1 cm.** showed significant differences at the levels of 0.01.

Figure 1 .
Figure 1.Phenotypic features of pod and seed weight of both JH5 and M130.(A) HPW of JH5.(B) HPW of M130.(C) t test for HPW of JH5 and M130.(D) HSW of JH5.(E) HSW of M130.(F) t test for HSW of JH5 and M130.Scale bar was 1 cm.** showed significant differences at the levels of 0.01.

Figure 2 .
Figure 2. Frequency distribution of phenotype data of HPW and HSW in RIL population under different environments.Gray rectangle displays HPW.Black rectangle displays HSW.The x-axis indicates the values of HPW or HSW in seven environments (2017QY, 2017DM, 2018QY, 2018DM, 2018QA, 2019XL and 2020QY).The y-axis shows the number of individuals in the RIL population.

Figure 2 .
Figure 2. Frequency distribution of phenotype data of HPW and HSW in RIL population under different environments.Gray rectangle displays HPW.Black rectangle displays HSW.The x-axis indicates the values of HPW or HSW in seven environments (2017QY, 2017DM, 2018QY, 2018DM, 2018QA, 2019XL and 2020QY).The y-axis shows the number of individuals in the RIL population.

Figure 2 .
Figure 2. Frequency distribution of phenotype data of HPW and HSW in RIL population under different environments.Gray rectangle displays HPW.Black rectangle displays HSW.The x-axis indicates the values of HPW or HSW in seven environments (2017QY, 2017DM, 2018QY, 2018DM, 2018QA, 2019XL and 2020QY).The y-axis shows the number of individuals in the RIL population.

Figure 3 .
Figure 3. Box plots of HSW (left) and HPW (right) of RIL population under different environments.

Figure 3 .
Figure 3. Box plots of HSW (left) and HPW (right) of RIL population under different environments.

Figure 4 .
Figure 4. IHDGM of the RIL population.Left ruler was the length of linkage group.Black indicator displays position of each marker on IHDGM.

Figure 5 .
Figure 5.The distribution of QTLs for HPW (blue indicator) and HSW (pink indicator) on the genetic linkage map.

Figure 6 .
Figure 6.Candidate genes for the co-localization interval of HPW and HSW.Co-localization interval LOD mapped on A08.Candidate interval corresponds to the physical position of chromosome A08.Distribution of candidate genes in the co-localization interval on A08.

Figure 6 .
Figure 6.Candidate genes for the co-localization interval of HPW and HSW.Co-localization interval LOD mapped on A08.Candidate interval corresponds to the physical position of chromosome A08.Distribution of candidate genes in the co-localization interval on A08.

Table 1 .
Descriptive statistics of HPW and HSW for parents and RIL populations.

Table 2 .
Simple correlation coefficients between HPW and HSW under seven environments.

Table 2 .
Simple correlation coefficients between HPW and HSW under seven environments.

Table 3 .
Results of ANVOA and h B 2 of HPW and HSW.Env. and G × E are abbreviation of genotype, environment and G × E interaction.
Figure 4. IHDGM of the RIL population.Left ruler was the length of linkage group.Black indicator displays position of each marker on IHDGM.

Table 5 .
QTL mapping results of HPW and HSW.

Table 6 .
Gene annotation in candidate regions.