| Rice (Oryza sativa) is one of the most important crops that feeds more than half people in the world. However, rice yield is easily affected by a number of diseases among which bacterial blight (BB), caused by Xanthomonas oryzae pv. oryzae (Xoo), is one of the three most serious diseases in China. Breeding practice has proven that development and application of of resistant cultivars with resistance (R) genes is the most economic and effective approach to control this disease.Three genetically interconnected multi-parents advanced generation inter-cross (MAGIC), including two populations (DC1 and DC2) derived from four parents and one population from eight parents (8way) were used to identify resistance QTLs to BB. Two weak virulent strains, C2 and C4, and two strong virulent strains, C5 and GD-V were employed to inoculate different MAGIC populations in this study. We also screened resistant breeding materials from the 3 MAGIC populations and estimate the prediction accuracy of BB resistance based on selectedwhole genomic markers. Since MAGIC populations bridage the genetic research and breeding program, they are of great significance in both genetic research and genomic selection in breeding program. The main results in this study were summarized as follows:1. The DC1, DC2 and 8way populations were not found containing significant population structure. In joint anlysis, the DC 12 population, combined by DC1 and DC2, and MAGIC Plus population, combined by DC1, DC2 and 8way, were found contain 2 and 3 subgroups, respectively. The result of LD analysis showed that LD declined to half of its initial value at about 2.5Mb in DC1 and DC2 populations, while at about 1.25Mb in 8way, DC 12 and MAGIC Plus populations.2. Phenotypic evaluation results showed that most parents were resistant to C2 and C4 but susceptible to C5 and GD-V. Transgressive segregations of lesion length for the four strains were observed in the three MAGIC populations and showed continuous distributions.3. A total of 15 QTLs affecting BB lesion length of four strains were detected. Most QTLs showed obviously genetic background effects, which means their resistance could be changed or even lost when transfer to different genetic backgrounds. This indicates that the background effects should be throughfully considered when these QTLs were used in breeding programs against BB. Among the 15 QTLs detected, QBbr11-1 and QBbr11-2 showed less genetic background effect, which showed stable contribution to BB resistance in both single population and joint populations association anaysis, demonstrating its high value in both genetic research and breeding program.4. Eight resistant lines with pyramiding of different genes/QTLs combinations were screened from the three MAGIC populations, and we found that combination of qualitative resistance gene with QTLs could significantly improve BB resistance. The eight resistant breeding lines could be used as important resistance donors for rice breeding against BB. Some crosses between resistant lines and susceptible lines selected from the 3 MAGIC populations were developed, which laid a solid foundation for fine mapping of major BB resistance QTLs.5. We estimated the prediction accuracies of whole genomic selected markers for resistances to C2 and GD-V strains, using DC1 and DC2 as training population while 8-way as validation population. When the single train population was used, the prediction accuracy for resistance to C2 is significantly higher than that for resistance to GD-V; while use joint training population could obtain best prediction accuracy for resistance to GD-V. For C2 resistance to strain, no difference on prediction accuracy was found between joint training population and single training population. |
PDF Full Text Request