Genome-Wide Association Study Of Important Agronomic Traits In Watermelon

Watermelon is an important fruit in the worldwide,and it is a very important horticultural crop in China which accounts for more than 60 % of the world's total productions.It is an annual herbaceous plant of Cucurbit crops.In this study,a total of173 watermelon accessions were used,of which 158 were collected by the USDA-ARS and 15 were collected by our research group.246 SSR markers selected from whole genome were used for genotyping in these watermelon accessions.Furthermore,26 phenotypic traits(cotyledon,vine,flower,fruit,seed and so on)of were phenotyped and analyzed by correlation analysis and principal component analysis.A mixed linear model(MLM)was used to correlate the phenotypic traits and genotypic data to identify SSR markers that were significantly associated with the target traits.The results will lay a foundation for further excavating the elite genes of important agronomic characters of watermelon and accelerating the molecular breeding process of watermelon.The detail results are as follows:1.246 SSR markers were selected for genotyping in 173 waermelon accessions,and Popgen32 were used to analyze the diversity bands detected from these markers.The number of observed alleles(Na)at each locus range from 2 to 11 with an average of4.32,effective allele frequency(Ne)range from 1.05 to 5.99 with an average of 2.08,observed heterozygosity(Ho)range of observation was 0.01-1.00 with an average of0.88,Shannon's information index (?) ranged from 0.11-1.98 with an average of 0.86,polymorphism information index(PIC)ranged from 0.04-0.81 with an average of0.42.2.Genetic diversity and population structure of 173 watermelon accessions were analyzed based on the fingerprinting of 246 SSR markers.The phylogentic tree showed that 173 watermelon materials could be divided into two groups: ?(Pop ?)included PI432337 and PI537300,and ?(Pop ?)included the remaining 171 watermelon accessions.Principal component analysis(PCA)results showed that the cumulative contribution of principal components was 14.4%,and these 173 watermelon accessions were divided into two categories.Area A included 10 watermelon accessions and area B included 163 watermelon accessions.Furthermore,we used a model-based approach for populationstructure analysis to analyze the 173 accessions.According to distribution of ?K values,there was only one peak of ?K when K=4 suggesting these watermelon accessions were grouping into four groups.3.The statistical descriptions and analysis of variance of 26 watermelon phenotypic traits for 173 watermelon test materials were performed using SPSS.The frequency distributions of 26 phenotypic traits were shown in a normal distribution.The correlation analysis results showed that there were extremely significant positive and negative correlations between the 26 phenotypic traits,the 17 traits with extremely significant positive correlation with the length of the pedicel,the 2 traits with being the most significant negatively correlated other traits.The results of principal components analysis showed that 26 phenotypic traits could be transformed into 8principal components,and the cumulative contribution rate of 8 principal components was 71.97 %.The results of variation analysis showed that the coefficient of variation of soluble solids content of margin pulp was the largest among the 26 traits(70 %),and the variation coefficient of the seed shape index was the smallest(0.08 %).Shannon Wiener index9(H')ranged from 1.32 to 1.67.The generalized heritability analysis results showed that the broad sense heritability of seed length and seed width was the largest suggesting it was less affected by the environment.4.The GWAS was carried out between 26 phenotypes and SSR genotypes using the MLM model by TASSEL 4.3.6.111 and 151 significant loci were identified for these traits in 2016 and 2017,respectively.Among of them,11 loci were detected in both environments.One locus was detected with fruit length,fruit width,flesh thickness,soluble solid content of seed pulp,seed width,melon pitch,and first male open flowering node,respectively,and four loci were associated with seed length.