First, confirm that you have the correct template for your paper size. This template has been tailored for output on the custom paper size (21 cm * 28.5 cm). The two connected populations used in the study consisted of 208 and 212 F10 RILs derived by single-seed descent from two crosses of Yu82 × Shen137 and Yu537A × Shen137, which were referred as Population 1 (Pop 1) and Population 2 (Pop 2) and used to identify QTLs for related trait of seed vigor, respectively.

The artificial aging treatment was used the same method described by Zeng et al. [26] . The seeds of two populations and three parents were reproduced in the winter in Hainan Province in 2015. After harvest, the seeds were fully dried under natural conditions. The seeds of each genotype were divided into four portions (60 seeds choosing to ensure sowing quality of every portion) for artificial aging treatments. All the seeds were placed in Nylon mesh belt firstly, then were treated at 45˚C ± 1˚C and 90% relative humidity for 0, 2, 4, and 6 days (0d, 2d, 4d and 6d) by using a thermostatic moisture regulator, respectively. Every treatment followed a randomized complete block design with three replications. Among treatments, 0d treatment acted as control.

The template is used to format your paper and style the text. All margins, column widths, line spaces, and text fonts are prescribed; please do not alter them. You may note peculiarities. For example, the head margin in this template measures proportionately more than is customary. This measurement and others are deliberate, using specifications that anticipate your paper as one part of the entire journals, and not as an independent document. Please do not revise any of the current designations.

The germination experiment conducted at 25˚C in artificial climate chamber in 2016. The method of germination experiment was as follows: the first, selecting diameter of 0.05 - 0.2 mm of fine sand as sprout bed and the sand was treated by high-handed sterilization pan at 120˚C for two hours; the second, using a germination container of 16 × 8 holes that the diameter of each hole was 40 mm; the third, each hole was filled with 3.5 cm thick sand and put 2 seeds in it, and then used 1.5 cm thick sand to cover them; the last, the germination containers sowing seeds were left in artificial climate chamber for a temperature 25˚C, a relative humidity 65% and illumination conditions 4000 lx, the photoperiod was 14/10 (day/night). The number of germinated seeds was counted daily. The data of related traits for 8 days after sowing were used for QTL analysis when obvious differences between the parents were observed. After daily statistics finished, 5 plants of each RIL were selected randomly to measure the seedling length, respectively. The germination percentage (GP) was calculated as GP = n/N × 100%, where n is the total number of germination seeds, N is the total number of seeds. The germination index (GI) was calculated as: GI = ΣGt/Dt, where the Dt is the germination time, Gt is the number of germinated seeds on the time. The vigor index (VI) was calculated as: VI = GI × SL, where the SL is the seedling length on day 8. The simple vigor index (SVI) was calculated as: SVI = GP × SL. The mean germination time (MGT) was calculated as: MGT = ΣGt × Dt/GP, where the sense of Gt and Dt as above.

The trait values for each RIL were reported as the average from five plants in each replication. The overall performance was the average over the three replications from each artificial aging treatment. Analysis of variance (ANOVA) was carried out to estimate genetic variation for all the measured traits among the RILs using the general linear model procedure of the statistical software SPSS 17.0. Descriptive statistics and simple correlation coefficients (r) between the traits were calculated using the above statistical software.

A total of 3072 pairs of single nucleotide polymorphism (SNP) markers were selected from the more than 800,000 SNPs to genotype the 420 RILs and three parents. We analyzed polymorphisms of 3072 SNP markers between two pairs of parents, Yu82/Shen137 and Yu537A/Shen137. Ultimately, 1397 and 1371 SNP markers had polyphisms between the two parents, respectively. Chi-square values were generated for 2768 SNP markers, 225 and 232 SNP markers showed serious segregation distortion and failed to be assigned to any linkage in the two populations. The linkage analysis was done with JoinMap version 4.0. Two genetic linkage maps were constructed with 1172 and 1139 SNP markers using Joinmap version 4.0 [27] , and the total length 1629.61 cM with an average interval of 1.39 cM for Pop.1 and 1681.75 cM with an average interval of 1.48 cM for Pop.2 [28] .

QTL analysis was conducted using composite interval mapping (CIM) with WinQTLcart 2.5 software [29] . For CIM, Model 6 of the Zmapqtl dodule was employed for detecting QTL and their effects, specifying the five markers identified by stepwise regression that explained most of the variation for a given trait as forward and backward parameters and a window size of 10 cM on either side of the markers flanking the test site [30] . To identify an accurate significance threshold for each trait, an empirical threshold was determined by performing 1000 random permutations [31] . QTL position was assigned to relevant region at the point of the maximum likelihood odds ratio (LOD). QTL confidence interval was calculated by subtracting one LOD unit on each side from the maximum LOD position [32] .

For the additive effects of QTL, positive and negative values indicated that alleles from the normal maize inbred lines Yu82/Yu537A and the maize inbred line Shen137 increased the trait scores, respectively. QTL were named according to ‘‘q’’ +‘‘artificial aging treatment days” + ‘‘trait abbreviation’’ + ‘‘population code’’ + ‘‘−’’ + ‘‘chromosome number’’ + ‘‘QTL number’’.

To integrate QTLs information for the measured traits located in the two connected RIL populations, the genetic linkage maps were integrated and consensus QTLs were identified by meta-analysis [33] [34] . The QTLs mapped in the two connected RIL populations were projected on the integrated map using their positions and confidence intervals shared by two linkage maps. Some controversial markers between two linkage maps were deleted, which could effectively improve the accuracy of projection.

Meta-analysis was performed by using BioMercator2.1 software [34] . The Akaike Information Criterion (AIC) was used to select the QTL model on each chromosome [35] . According to this, the QTL model with the lowest AIC value is considered a significant model indicating the number of meta-QTL. The number of mQTLs that best fitted the results on a given linkage group was determined based on a modified Akaike criterion [36] . Meta-QTL were named according to ‘‘q’’ +‘‘artificial aging treatment days” + ‘‘trait abbreviation’’ + ‘‘population code’’ + ‘‘−’’ + ‘‘chromosome number’’ + ‘‘QTL number’’.

The values of GI, VI and SVI were obviously decreased and the values of MGT were markedly increased after three treatment conditions compared with control in parents and two populations. For three parents, the values of GI, VI and SVI were higher for Yu82 and Yu537A than Shen137 under four aging treatments, while the reverse was true for MGT. trait differences were also found among three parents under each treatment. For RILs, the values presented a large range of variability with transgressive segregation exceeding values of high values parent. All traits showed normal distribution in the two RIL populations and differed substantially under various treatment conditions (Table 1).

Significant positive correlations were consistently observed for GI, VI and SVI from two RIL populations under control and after various aging treatments except for between VI and SVI from Pop. 1 under 4d aging treatment, while MGT and GI, VI, SVI showed significant negative correlations except for between MGT and SVI from Pop. 2 under 0 and 2d aging treatments (Table 2).

A total of 74 QTLs for GI, VI, SVI and MGT were detected in two connected populations under control and after three aging treatment conditions, with 40 QTLs in Pop. 1 and 34 QTL in Pop. 2 (Table 3). These QTLs were located on all chromosomes. The contributions to phenotypic variations for a single QTL ranged from 5.33% to 13.74%, with 10 QTLs over 10% and 1 QTL over 13%.

GI

Nine QTLs in Pop.1 and ten QTLs in Pop.2 were identified and located on all chromosomes except for chromosomes 2 and 10 under four aging treatment conditions. The contribution rates of these QTLs ranged from 5.56% to 13.74% of total phenotypic variance (Table 3). The positive alleles of q2GI1-3, q2GI1-5-1, q2GI1-5-2, q4GI2-4 and q6GI2-9 were derived from Shen137 to contribute towards an increase in values of GI. There were qGI1-6 from Pop.1 consistently mapped in the same marker interval SYN31854-PZE-106102131 after 2d and 4d aging treatments, qGI2-8-2 from Pop.2 in the interval PZB00865.2-PZE-108073195 after 0d and 2d aging treatments, and qGI2-8-1 from Pop.2 in the interval PZE-105077135-PZE-105082252 after 0d, 2d and 4d aging treatments. Among these QTLs, QTL qGI2-8-1 was responsible for 10.25, 11.73 and 8.22% of phenotypic variance, and qGI2-8-2 responsible for 10.82 and 7.07% of phenotypic variance, respectively.

VI

Eighteen QTLs were mapped for VI under control and after three aging treatments in the two populations, nine in Pop.1 and nine in Pop.2. They were distributed across the whole genome, except for chromosomes 9 and 10 with contribution to phenotypic variation for a single QTL from 5.39 to 8.89% (Table 3). The positive alleles of qnVI1-1-1, q2VI1-1, q4VI1-1, q4VI21-6 and q6VI1-7 in Pop.1 and of qnVI2-8 in Pop.2 were contributed by Yu82/Yu537A. However, there was no QTL identified at same marker intervals under different aging treatment conditions in the two populations.

SVI

Eleven QTLs for SVI were detected, with four in Pop.1 and seven in Pop.2 under all aging treatment environments. They were distributed across chromosomes 2, 3, 4 5 and 6, and explained 5.72 to 12.11% of the phenotypic variation (Table 3). Among these QTLs, two in Pop.1 and one in Pop.2 were derived from Yu82/Yu537A to increase in the trait values. The positive alleles of qSVI2-2