QTLs for oil content and their relationships to other agronomic traits in an European x Chinese oilseed rape population

Abstract

Rapeseed oil is one of the most important vegetable oils for edible consumption both in China and Europe, while the meal, after extraction of the oil, has a high protein content with well balanced amino acid composition and is a valuable protein supplement in animal feed mixtures today. Therefore a better understanding of the genetic basis for oil and protein content by mapping QTLs ( Quantitative trait loci ) with molecular markers is important for rapeseed breeding. Analyzing the genetic relationships of these two traits with developmental stages, and with yield related traits on QTL level is of interest. A total of 282 doubled haploid ( DH ) population, which derived from the F1 between the old German cultivar Sollux and the Chinese landrace Gaoyou of Brassica napus by microspore culture, was developed and a linkage map including 125 SSR marker loci was constructed. The DH lines were grown in 4 environments, two in Germany ( Reinshof and Weende ) and two in China ( Hangzhou and Xian ). A new mapping program QTLMapper version 1.0 was used for mapping unconditional and conditional QTLs with additive (a ) and additive × additive epistatic ( aa ) effects, as well as their interaction effects with environments ( QE, including ae and aae ). Both parents showed high oil content. Although the maximum difference between parents either for oil or for protein content were around 3 % in all locations, large transgressive segregation for both traits within the DH lines was observed in each environment with extreme differences up to 17 and 10 percentages for oil and protein content in Germany, and 10 and 7 %, respectively, in China. Mapping results revealed 8, 5 and 7 QTLs with additive main effects, and 9, 6 and 6 pairs of digenic loci with epistatic main effects for oil, seed and meal protein content, respectively. The additive effects of these QTLs explained about 40, 30 and 40 %, and the epistatic effects of theses pairs accounted for about 40, 25 and 30 % of the phenotypic variation observed for oil, seed and meal protein, respectively. For oil content, two important alleles were identified with Chinese parent and two alleles in the European parent. These QTLs were confirmed by the comparison of the 20 highest and 20 lowest DH lines. A large number of QTLs with ae interaction effects demonstrated the complex inheritance for oil and protein content and suggest a joint utilization for QTLs with additive, epistatic and QE interaction effects by marker assisted selection in breeding programs. By combining the statistical procedures for analyzing conditional genetic effects and the QTL mapping program, conditional mapping analysis between oil and seed protein were performed. Results of conditional mapping present the effects of QTLs for oil content under the condition of no variation for seed protein content, and vice versa. Comparison between unconditional ( conventional QTL mapping) and conditional (oil conditioned on seed protein: oil|seed protein, or vice versa ) mapping results, it is possible to find the genetic relationship between two traits on individual loci. The results can be summarized as following points 1) to distinguish the pleiotropic effect or gene linkage in the same genomic regions. Two QTLs might be regarded as pleiotropy because under conditional mapping for oil|seed protein and seed protein|oil, the QTLs did no longer show significant effects for both traits, and one QTL more like gene linkage since reduced genetic effects were detected for both traits under conditional mapping one another. 2) two QTLs for oil and seed protein content respectively were proved to be independent while three oil QTLs were significantly influenced by seed protein content. 3 ) large changes of QTLs with epistatic effects under conditional mapping analysis reflected the close genetic relationship between these two traits also through digenic epistatic loci. Oil and protein content in the seed share large part of their genetic basis. However, it is possible to combine high content of seed oil with high meal protein content. Four of seven additive QTL and two of six pairs of epistatic loci for meal protein were not affected by oil content and thus seed oil and meal protein show much less genetic overlapping than seed oil and seed protein. The genetic influence on QTLs for oil and protein content due to the variation of developmental stages were analyzed by conditional mapping. Flowering time had more contribution to oil content, while flowering period and grain filling period were more related to seed and meal protein. The two oil QTLs having largest effects were identified independently of all developmental stages. The favorable alleles come from the Chinese parent. One QTL with favorable alleles from the European parent for oil content also did not relate to any developmental stages. Conditional mapping for oil and protein content when seed weight, seed number and plant height were conditioned, revealed that oil and seed protein were influenced by seed number and plant height by both additive and epistatic effects. The genetic relationship between oil, seed protein and seed weight were small. Seed size and seed number are important yield components, and silique length is normally highly correlated with seed number. In total, 7, 6 and 9 QTLs respectively, were identified for these traits which explained about 40, 35 and 70 % of the mean phenotypic variation, respectively. Digenic epistasis was small for seed size, negligible for silique length but significant for seed number. QTLs for seed size were more related with flowering time and grain filling period, while QTLs for seed number were more influenced by flowering period. In the last part of this thesis, The QTLs for two plant type traits: plant height and number of primary branches, and two important developmental traits: flowering time and maturity were mapped. The results demonstrated that each trait was controlled by several additive main QTLs plus a number of QTLs with epistatic and QE interaction effects. Plant height was controlled by many QTLs ( 12 loci with a or combined ae, 5 loci only with ae ). Seven of eight main QTLs showed early flowering alleles from Gaoyou and no clear transgressive segregation was observed among DH lines. It is possible to create new breeding materials or cultivars with increased oil content in seed and protein content in meal by marker assistant integration of favorable genes identified in this study. The genetic relationships between QTLs for oil ( or protein ) content and developmental stages ( or yield related traits ) further provide useful information for the utilization of these QTLs in both regions.