Genetic Analysis, QTL Mapping And Association Mapping For Panicle Traits In Japonica Rice (Oryza Sativa L.)

Japonica rice growing area in China is 7,011,000 hm2 each year and accounts for more than 50 percent of world’s japonica area (12,542,000 hm2). The area planted with japonica hybrid rice only occupied 3% of the total area of japonica rice in China. Therefore, great space exists for developing japonica hybrid rice, compared with indica hybrid rice, in which great achievement had been made. Japonica hybrid rice has advantage in resistant to disease and pests, but competition heterosis in yield is not obvious. Improving panicle traits is an efficiency method for yield enhancement. Discovering favorable alleles for panicle traits is the foundation for panicle trait improvement. In this study, genetic segregation analysis of primary branch number and secondary branch number per panicle were performed firstly by using 370 recombinant inbred lines (RILs) and their parents, Xiushui 79 and C Bao, which are both japonica rice cultivars. Secondly, a SSR linkage map in japonica rice was constructed by using 254 RILs selected from above 370 RILs. Thirdly, QTLs of panicle traits and flag leaf traits were detected based on the map. Finally, association analysis between panicle traits and SSR markers was conducted in natural population composed of 58 core collections of landrace and 36 cultivars currently used widely in japonica rice in Taihu Lake valley. The main results are as follows:1. Primary branch number per panicle was controlled by two linkage major genes plus polygenes. Major gene heritability was 66% in 2005 and 40% in 2006, and polygene heritability was 11% in 2005 and 41% in 2006. Secondary branch number per panicle was controlled by two major genes plus polygenes. Major gene heritability was 80% in 2005 and 65% in 2006, and polygene heritability was 10% in 2005 and 12% in 2006. The results above were obtained by analyzing 2 years of segregation data of primary branch number and secondary branch number in RIL population (370 lines, F7:8 in 2005 and F8:9 in 2006, Xiubao RIL population for short)) made from Xiushui 79 and C Bao with the mixed major-gene plus polygenes inheritance model.2. Among 915 pairs of SSR primers used for screening polymorphism between Xiushui 79 and C Bao,105 pairs of primers amplified polymorphic products using total DNA of the two parents as template. Chromosome 1 showed the smallest polymorphism rate (7.8%). Chromosome 7 showed the largest polymorphism rate (16.4%). A SSR linkage map was constructed by genotyping 254 family lines (F10:11) from Xiubao RIL population by using above 105 SSR markers. The map containing 91 information loci has a total distance of 969 cM, averaging 10.6 cM between two loci. The total allele frequency for the RIL population was calculated at 0.537 and 0.463 for Xiushui 79 and C Bao, respectively, fitting to the expected allelic frequency of 1:1.3. Phenotypes values of six panicle traits and four flag leaf traits in two growing environments were investigated in 254 Xiubao RILs and their parents. QTLs of additive effects and additive×additive effects for these ten traits were detected and their interaction with environments was analyzed based on above SSR linkage map. The results showed that:(1) 53 additive effects QTLs for all ten traits were identified. Five additive QTLs were detected for primary branch number per panicle, explained 4%～11% of the phenotypic variation; Four additive QTLs were detected for secondary branch number per panicle explained 2%～16% of the phenotypic variation; Six additive QTLs were detected for panicle length, explained 1%～34% of the phenotypic variation; Six additive QTLs were detected for spikelet number per panicle, explained 2%～11% of the phenotypic variation; Six additive QTLs were detected for filled grain number per panicle, explained 3%～6% of the phenotypic variation; Four additive QTLs were detected for spikelet density, explained 2%～21% of the phenotypic variation; Three additive QTLs were detected for flag leaf length, explained 8%～40% of the phenotypic variation; Five additive QTLs were detected for flag leaf width, explained 4%～16% of the phenotypic variation; Five additive QTLs were detected for flag leaf area, explained 0.6%～14% of the phenotypic variation; Nine additive QTLs were detected for leaf rolled index, explained 2%～9% of the phenotypic variation. (2) Three pleiotropic effects loci were found which have simultaneous effects on flag leaf traits and panicle traits, two large effect loci in RM6570-RM5652 and RM5652-RM410 on chromosome 9, respectively, have simultaneous positive effects on FLL and PL. They explained 41% and 40% of the phenotypic variation in FLL, respectively,34% and 34% of the phenotypic variation in PL, respectively. The pleiotropic was confirmed by correlation analysis that FLL and PL have the largest contributions to each other (r=0.86, P< 0.01). (3) Twenty nine epistatic interaction QTL pairs detected for all traits except SBN, explained 1.0%～17.5% of the phenotypic variation for single trait.4. The genotyping data of 91 SSR markers on representative sample of 58 core collections of Japonica rice landrace in Taihu Lake valley and 36 accessions currently used widely in japonica rice production were used in the present study. Linkage disequilibriums of pairwise loci were analyzed for the two populations and population structure was analyzed for the natural population which was composed of the two above populations. Then the association analysis between SSR loci and 7 panicle traits was performed by using TASSEL GLM (general linear model) program. Alleles of loci significantly associated with the traits in two environments were analysised. The results showed that:(1) Various degrees of LD were detected not only among markers on the same chromosomes but also among markers on the different chromosomes, and the LD attenuation was slow. (2) Genetic structure analysis showed that the natural population was composed of eight subpopulations, which associated with their heading date eco-types, indicating the classification of heading date eco-types was of found genetic bases. (3) Twenty eight loci associated with the panicle traits were screened out from the natural population. Seventeen (60%) loci were found to associate with two or more traits simultaneously, which was confirmed by correlation analysis that most traits were significant correlated to each other. (4) Fifty favorable alleles and 28 their carrier materials were screened out.