Increase of seed oil content in winter oilseed rape (Brassica napus L.) by using Chinese genetic resources

Oilseed rape (Brassica napus L.) is one of the world’s most important oil crops, and due to a growing demand of vegetable oil for nutritional as well as industrial purposes increasing seed oil content is a major aim for oilseed rape breeding. Compared to other field crops oilseed rape is a rather young species and therefore its genetic variation is limited. However, breeding efforts of the last decades increased seed oil content substantially in European breeding material. But the intensive selection process further decreased genetic diversity, thus the need for a new source of genetic variation rose. As a candidate of high potential to lead to a rapid breeding success, Chinese breeding material, which independently underwent an intensive selection for high yield and oil content during the last decades, was chosen in a previous study. Within this, a quantitative genetic analysis was conducted on a DH population derived by a cross between the Chinese cultivar Gaoyou and the European cultivar Sollux, both with high oil content, but also high erucic acid and glucosinolate content. Thereby, the DH line 14 (SGDH14) was identified as best performing line under European conditions with highest oil contents and a combination of all favourable QTL alleles for oil content from both parental cultivars. Thereupon SGDH14 was crossed to Express617, an inbred line of Express, a European high oil cultivar of canola quality, since erucic acid and glucosinolates are undesirable traits in modern breeding material. And a F1 derived DH population (SGEDH) consisting of 212 genotypes was developed segregating for erucic acid and glucosinolate content. In the present study, the SGEDH population was analysed to find out more about the genetic variation and inheritance of seed oil content, the underlying fatty acid composition and other seed quality traits, and to identify new QTL responsible to further increase oil content. Thus, field experiments of the SGEDH population were conducted in the two mega-environments, Europe (North Germany and South Sweden; EU trials) in three consecutive growing seasons 2009/10, 2010/11 and 2011/12, and East China (Hangzhou; Chinese trials) in two consecutive growing seasons 2010/11 and 2011/12. And a genetic map was constructed including 15380 SNP, 314 DArT and 116 AFLP markers, organized in 19 linkage groups and covering 2651cM. This map comprised 1693 individual marker positions. QTL mapping was conducted using a framework map consisting of a subset of 379 markers selected form the full map, and applying the composite interval mapping (CIM) approach of QTLNetwork software version 2.1. Mega-environments were investigated separately, and results subsequently compared. Analysis of variances of the SGEDH population revealed highly significant genotypic effects for all traits in both environments, except for flowering period in the Chinese trials. The phenotypic variation was moderate to high depending on the trait considered. Seed oil content in the SGEDH population ranged from42.5 to 50.9% in the EU trials, and from 39.4 to 49.8% in the Chinese trials, showing a slightly broader range for seed oil in the Chinese trials, but in total higher oil contents in the EU trials. Heritability for oil content was high as well for the EU trials with 0.96 as for the Chinese trials with 0.94. Heritabilities for all other traits investigated in the EU tails were high ranging from 0.76 to 1.00. High heritabilities ranging from 0.70 to 0.99 were also found for most other traits in the Chinese tails, except for end of flowering (0.55) and flowering period (0.1). High significant positive correlations (P = 0.01) were consistently observed in both trials between oil content and erucic acid content, protein content in defatted meal and plant height. Since a strong association between oil content and erucic acid was reported previously and confirmed by a high positive correlation between these two traits in this study, corrections of oil content were conducted to eliminate the effect of erucic acid on oil content. This enabled a comparison of the oil contents of genotypes with varying erucic acid contents, and identified SGEDH175 and 13, originally of medium erucic acid content, as the genotypes with highest oil contents in the EU trials, and SGEDH210 and 145 in Chinese trials. Corrected oil contents were subsequently used to identify additional QTL for oil content independent of erucic acid content. In the group of erucic acid free genotypes SGEDH172 was identified with highest oil content in both mega-environments. QTL mapping applying the CIM method identified four QTL for oil content in the EU trials and six in the Chinese trials. Individual QTL explained between 10 and 50.5% of the phenotypic variance in the EU trials and between 3.3 and 46.5% in the Chinese trials. Four additional oil-QTL were found for corrected oil contents in the EU trials. Comparison of all QTL results for oil content from both mega-environments, including QTL for corrected oil contents, revealed that three QTL were constantly detected. These three environmentally stable QTL were located in overlapping genetic regions on linkage groups A08, C03 and C05. Furthermore, four environmental-specific QTL for oil content were identified in EU trials, located on A10, C03, C04 and C05, while three individual oil-QTL were found in Chinese trials on A06, A07 and A10. For other traits, between two and eight QTL were identified in the EU trials, and between two to five QTL were identified in the Chinese trials. Individual QTL explained between 0.6 and 64.1% of the phenotypic variance in the EU trials and 0.0 and 63.8% in the Chinese trials. Comparison of QTL results revealed that between one and three QTL were repeatedly identified for other traits across the two trials. BLAST search of SNP and DArT marker sequences to the B. napus Darmor-bzh reference genome enabled the identification of the physical position of 1289 SNP and 94 DArT markers from the set of 1693 individual full map marker positions. Alignment of genetic and physical map positions of markers showed a good collinearity for all linkage groups, except for linkage group C09. Physical positions of putative candidate genes involved in storage oil biosynthesis were compared to the physical positions of markers within QTL confidence intervals for seed oil content. This identified two genes co-located with QTL for oil content in the EU trials and two genes co-located in the Chinese trials. FAD3, encoding the linoleic acid desaturase, was located within the confidence interval of the oil-QTL on C04 in EU trials, and KAR, encoding the ketoacyl-ACP reductase an enzyme of the fatty acid synthase complex, was located within the confidence interval of the oil-QTL on A06 in Chinese trials. In both trials FAE1, encoding a ketoacyl-CoA synthase involved in the elongation of oleic acid to erucic acid, was found co-located with the stable oil-QTL identified on A08. Of highest interest for further breeding the major oil-QTL for corrected oil content E_Oil-reg_corr-3 on C05 was identified in the present study.