| Since rice panicle traits directly relate to rice yield and quality, much emphasis have been place on panicle study. So far, there have been a large number of reports on panicle traits, such as panicle length, primary branch number, second branch number, grain number per panicle, panicle angle and so on. And many genes have been cloned. Because of the panicle dispersity and low yield of wild rice, compact panicle was chosen as a favorable trait during rice domestication. Wild rice spices have sparse panicle branches and scatter grains; domesticated rice cultivars have more branches and higher grain density, and panicles bend at maturity. During indica rice cultivars'expansion to template regions where are not suitable for their growth, new cultivars that were more resistant to low temperature were bred. To improve rice yield, panicle type were developed from bending panicles to erect ones, and become more compacted. However, compact panicles could lead easily to the outburst of rice false smut disease, which affect rice yield and quality directly. In the last decade, japonica rice cultivars with erect panicle have been planted widely in japonica rice growing regions, and at the same time, rice false smut disease has become more and more serious in those areas, especially in Northeast China. Thusly, resistance to rice smut disease has become a primary goal in japonica rice breeding in Northeast China. Therefore, regulating rice panicle dispersity will play a positive role in coordinating rice yield and quality with resistance to disease. But genetic researches on dispersity related traits are rarely reported.This study was carried out by using 254 recombinant inbred lines (RILs) and their parents, compact panicle type Xiushui79 and disperse panicle type C Bao, which are both japonica rice cultivars. Firstly, genetic segregation analysis of average primary branch angle (PBA) was conducted by using the mixed major gene plus polygene inheritance model. Secondly, QTLs were analyzed by softwares of QTLNetwork 2.0 and WinQTL Cartographer 2.5. All results were as follow: 1. In the RIL population, E-2-0 was determined as the best-fitted inheritance model for the primary branch angle trait in two growing environments, showing that this trait followed the mixed inheritance of two major genes with additive-epistatic effects plus polygene with additive-epistatic effects. The heritability of major genes were 65.49%(E1) and 52.69%(E2), while the heritability of polygenes was 2.43%(E1) and 32.46%(E2).2. Analysis based on multiple regression models shows that 21 QTLs for 4 traits are detected in two environments. qPBAⅠ9, qPBAⅡ9, qPBAⅢ9 and qPBA9 were detected stably explaining 14.44%～30.57% of phenotypic variations.3. Analysis using on mixed model based composite interval mapping demonstrated that 25 QTLs were deteced for primary branch angle. Among them 14 QTLs (qPBAl,qPBA6, qPBA9, qPBAⅠ1, qPBAⅠ3, qPBAⅠ6, qPBAⅠ9, qPBAⅡ1, qPBAⅡ4, qPBAⅡ6, qPBAⅡ9, qPBAⅢ1, qPBAⅢ5 and qPBAⅢ9) were detected in 2 methods. Moreover,2 epistatic interaction QTL pairs were also detected. Epistatic interaction between interval RM164-RM305 and interval RM5609-RM5479 increased primary branch angle, while epistatic interaction between interval RM8105-RM84 and interval RM448-RM8277 decreased it. No QTL-environmet effect was detected.4. QTLs located on the interval of RM3700-RM3600 in chromosome 9 (qPBAⅠ9, qPBAⅡ9, qPBAⅢ9 and qPBA9) that were stably detected in 2 environments by 2 methods explained most of the phenotypic variation and were reliable QTLs. |
PDF Full Text Request