Association Mapping Of Yield And Quality-related Traits In Brassica Napus L

Rapeseed（Brassica napus L.） is a widely cultivated oil crop in the world. Yield and quality-related traits in rapeseed are increasingly taken great importance. QTL mapping and biological metabolic pathway analysis were done to dissect the genetic bases of yield and quality-related traits in Brassica napus. Association mapping （AM） with its high resolution is a powerful approach in quantitative trait loci mapping, and has been widely adopted in complex genetic diseases in human and important agronomic traits in crops. In the study, a panel of192inbred lines of Brassica napus from all over the world was genotyped using451single-locus microsatellite markers and740amplified fragment length polymorphism （AFLP） markers. Genome-wide association studies were conducted with six yield-related traits and ten quality-related traits in our association mapping, main results are as follows:1. Phenotypic analysis of yield and quality-related traits in Brassica napus:In our association mapping, extensive phenotypic variations were observed for all traits, and all traits had an H2higher than80%, suggesting that all these traits are stably inherited. Highly genetic and phenotypic correlations were observed between almost all the traits. The phenotypic analysis suggesting that the phenotypic traits can be applied to the following association mapping.2. Model comparison on controlling false associations:ANOVA results indicated that most of all measured traits showed a strong influence of population structure （P<0.01）, and most of all traits were significantly different between subgroups （P<0.01）. These results indicating that population structure had significant impacts on phenotypic variation for almost all traits. The comparision with six statistical models in association analyses indicating that the Q+K model could effectively control false positive associations and was chosen for association analyses between the SSR and AFLP markers and yield and quality-related traits.3. Association mapping of six yield-related traits:Using the optimal Q+K model, association analyses were carried out for the six yield-related traits using the means across three years in Brassica napus. A total of37associated loci were detected, with two to fourteen markers associated with individual traits. The effects of phenotypic variance explained by associated markers were relatively small, ranged from3.51%to9.03%. And 12markers were located within or close to QTLs identified in previous studies. Among these, five markers （EA02MC04₇, EA14MC08J, EA09MC11）5, EA14MC02₇and EA03MC04₁0) were repeatedly associated with first branch height and plant height, and one common marker was detected association with silique length and seed weight.4. Association mapping of ten quality-related traits:Using the optimal Q+K model, association analyses were carried out with the369SSR and740AFLP markers for ten quality-related traits in Brassica napus. A total of177associated loci were detected in three years （P<0.001）, the effects of phenotypic variance explained by associated markers were ranged from0.11%to16.56%. Of these,79markers were repeatedly detected in two or three years. The correlation analysis demonstrated that highly genetic and phenotypic correlations were observed between almost all the traits. We found that61common markers were associated with correlated traits in our association mapping. Among these, two markers （EA06MG06-8and EA09MC08-7） were repeatedly associated with both oil content and protein content.32,30,31common markers were detected association with eicosenoic acid and oleic acid in three years, respectively, which was the most common markers associated with correlated traits. For ten quality-related traits in three years, some markers were repeatedly detected association with multiple traits in multiple years. For the marker EA06MG08₈, was significantly associated with erucic acid, seed glucosinolate, eicosenoic acid, oleic acid and palmitic acid in all three years, and was significantly associated with stearic acid in2009,2010.5. Association mapping of the principal components axis for ten quality-related traits: Using the PCA analysis, ten quality-related traits were synthesized into the first three principal components axes which totally accounting for86.47%,87.51%,87.53%of phenotypic variance in three years, respectively. The first PC axis accounted for erucic acid, seed glucosinolate and fatty acid composition; the second PC axis accounted for oil content and protein content; the third PC axis accounted for linolenic acids. And we conducted association analysis with these three PC axes.30associated markers detected with the first PC axis in two or three years and6associated markers detected in one year, were included in57common associated markers which detected with more than two traits in ten quality-related traits but oil content, protein content and linolenic acids. No common associated marker was detected with the second PC axis and oil content, protein content.3common associated markers detected with the third PC axis and linolenic acids （BnEMS620, EA04MG04₃and EA09MC14₁1）. 6. Allele effects of common associated markers on yield and quality-related traits: We found that for each associated marker detected with correlated traits, the lines contained one allele was preferentially shared higher content than that contained the other allele for positively correlated traits, and was shared lower content than that contained the other allele for the traits, which were nagetively correlated with positively correlated traits. For the marker EA06MG088, the lines contained the allele A0were preferentially shared higher content than that contained the allele A1with erucic acid and eicosenoic acid, and were shared lower content than that contained the allele A1with oleic acid. And the combination effects of haplotypes of associated markers were preferentially higher than the effects explained by the individual markers. EA02MC04₇and BrGMS2998explained6.67%and7.33%of phenotypic variance of plant height, respectively, while they jointly explained22.48%of phenotypic variance. These results suggested that the pyramiding of favorable alleles with minor effects is an effective way to enhance trait performance of rapeseed variety for yield and quality-related traits, which depict us the perspective of application in molecular breeding.