| The genetic basis of heterosis has been debated over a century. Using an F2:3 population derived from a cross between "Zhenshan 97" and "Minghui 63" , the parents of Shanyou 63 that is the most widely used rice hybrid in China, we previously showed that epistatic interactions play an important role as the genetic basis of heterosis (Yu et al 1997). However, compared with the F2 data, a major shortcoming associated with the analyses of the F2.3 data is an underestimation of dominant types of genetic components, such as dominance and overdominance at single locus level and dominance by dominance interaction at two-locus level. Another shortcoming associated with the data of all previous studies of similar subject is that the analyses were based on the performance measurements of the traits rather than heterosis. To overcome the shortcoming compared with the F2 3 data, we designed a new mating scheme that produced an "immortalized F2" population, using a population of 240 recombinant inbred lines (RILs) produced by single seed descendant method from the Zhenshan 97/Minghui 63 cross.In this design, three rounds of permutation resulted in 360 crosses. As many plants as one wish thus permitting replicated trials represent each of the genotypes in this population. So the field experiment can be repeated in multiple environments. Plants used for heterosis study are hybrids rather than progenies of self-fertilization. Comparing with parents, the heterosis for each cross could be evaluated. In this dissertation, we use the molecular markers to analysis the genetic model associated with 12 traits' performance and their heterosis. The main results are summarized as follow:1. We constructed the "immortalized F2" population that included 324 and 358 crosses in 1998 and 1999, respectively. The field experiments were completed in successive year. Randomized complete block designs with two replications have been conducted in 1998 and 1999. Each block containing four rows of 10 plants: two rows of the hybrid and one row for each of the parents. Hybrid plants of the crosses were identified by morphological comparison and SSR analysis. There are more than 90% of the crosses that had more than half individuals were true hybrid.2. As a result, 231 informative markers, screened for polymorphism with total 700 plus markers, were used for genotyping and mapping in the RI population. The molecular marker data need in the "immortalized population" could be collected only from the 240 RILs. The genome and their proportions are similar to those of an F2 population. The proportions of the genotypes in the "immortalizedF2" population, the genome of Minghui 63 and Zhenshan97 is 27.08% and 24.03%, and the heterzygousity frequency is 49.04%. The gene frequency of 117 loci skewed from 1: 1 segregation, of which 77 loci skewed towards Minghui 63 and 40 towards Zhenshan 97, while the number of the loci is 45 towards Minghui 63 and 23 towards Zhenshan 97 in F9 RI population, respectively.3. A genetic linkage map with 231 marker loci spans 2007.3 cM covering 12 rice chromosomes with an interval of 8.69 cM, which well integrated the markers from RGP and Cornell maps. The marker order on the map is similar to which of F2 population.4. Transgressive segregation was observed for all traits' performance and their heterosis in the "immortalized F2" population. There existed significant difference in all traits' performance between the two years. The interactions between genotype and environment were observed for all traits, and the environment effects on tillers per plant were relatively larger than than on the other traits.5. The number of QTLs greatly varied according to the different traits. 25 and 27 QTLs were detected for yield and three yield component traits in 1998 and 1999 respectively (37 QTLs totally), and 15 are common. For other traits, there were 64 QTLs detected. Of which 32 QTLs were simultaneously detected in two years. Majority of the QTLs explain small proportions of the variance for heterosis. We detected QTLs...
|
PDF Full Text Request