QTL Mapping For Some Vegetative Traits And Genetic Dissection Of Heterosis In Brassica Napus

Since the early day of last century, many theorists have given different explanations to heterosis, which is a very complicated phenomenon in most biology. Davenport (1908) and Bruce (1910) brought forward the dominance hypothesis, and almost at the same time Shull (1908) and East (1908) took out their over-dominance hypothesis. In 1990s', the advancement of QTL mapping method established the genetic basis to the mechanism of heterosis. Stuber et al (1992) first explored the connection of QTLs' interaction and heterosis in corn and thought the over-dominance was the genetic base of heterosis; Xiao et al (1995) researched the QTLs' activities and heterosis in rice, and said the dominance is the essence of heterosis; Yu et al (1997) summarized the their research outcomes of the rice hybrid Shanyou63 and put forward that the epistasis mostly contributed to the heterosis expression. All these theories based on analysis plenty of field data and macroscopically explained the quantitative traits by utilizing some QTLs' interaction, but they didn't concern the mutual affection of different traits in the developmental course and neglected the objective traits were controlled by many genes so that they couldn't understand the cause of heterosis at a deeper level.There are almost no reports on using Brssica napus populations to investigate heterosis, but many traits of rapeseed have been localized on its linkage groups today. The authors always use three major kinds of populations: F₂, BC₁ and DH but they all have some demerits. The F₂ population can't be set repetition, and BC₁, DH can not calculate the dominant and additive effect, respectively.For seek understanding for heterosis well, we choose the vegetative traits which have more MPH relatively and the immortalized F₂ population (IF₂) which can remedy the flaws of the three traditional ones. And we aim to offer a valuable consult for theorists and breeders in research on heterosis.In this study, 258 DH lines which derived from the Brssica napus combination Quantum2 × NO2717-17 were used to establish IF₂. Based on this population, a genetic linkage map for rapeseed was constructed with 112 SSR markers and 239 SRAP markers, and covered approximated 1717cM genome in length with an average interval of 5.3cM.It has 20 linkage groups (LGs). LG1 and LG10 are the longest ones which comprise 45 markers; LG14, LG16, LG18 only has 3-4 markers each, and other groups have 8-22 markers.We analyzed the data of four vegetative straits, and found all the four vegetative traits were quantitative traits. Fifteen QTLs were detected by ANOVA at the significant level P ^0.01 for four vegetative traits. They concentratively located on LG1, LG10 and LG19. Each of these QTLs can explain the phenotypic variation from 1.81% to 14.52%. We found 12 QTLs had part-dominant effects, and the other 3 had over-dominant effect. The higher the correlation coefficient is, the nearer the QTLs of two traits are, and even overlapped on the map.At the significant level of P<0.005 in two locations, we detected 177 loci pairs between co-dominant markers for the four traits by using Excel Macro. Interaction pairs all were non-QTL at the level of P<0.001. A great variety of interaction ways exist in the genome of four vegetative traits. One locus can affect different traits; one pair of interaction locus can also act with different effect to different traits. The coverage of interaction is broad, the locus involved were 10.46%-34.88% of all the co-dominant markers for different traits. We deemed epistasis interactions might play an importance role in of vegetativetraits expression in Brassica Napus.120 loci pairs were detected for the interaction of four traits' MPH at the level of P<0.01 by Excel Macro between co-dominant markers. Epistasis interaction mainly expressed by addictivity X dominance, and the markers concerned were more than 20% of all. So we thought epistasisis crucial in the genetics of the traits' MPH.