Development Of Different Types Of The Indica×Japonica Population And Genetic Mapping Of Important Quantitative Traits In Rice(Oryza Sativa L.)

Rice is one of the important crops in the world. To meet the ever-increasing demand for food from the increasing population, the major task and challenge are to improve the grain yield and quality under limited land resources and water supplies. One optimum ponetial approach to solve the problem is to use the hybridization between indica and japonica, i.e. to improve indica by using japoniaca genes, and in the meantime to improve japoniaca by using indica genes. In this study, we developed and improved six types of genetic population that were derived from Oryza sativa ssp. japonica cv. Asominori and Oryza sativa ssp. indica cv. IR24. The six populations were: one RIL population consisting of 215 recombinant inbred lines(RIL); two improved reciprocal sets of chromosome segments subtituion lines(CSSLs), one consisted of 64 IR24 CSSLs with Asominori as the genetic background(NAIS) and the other constsed of 57 Asominori CSSLs with IR24 as the genetic background(NIAS); two F1 populations, one(designated as HAIS) was made by crossing the NAIS population with Asominori, and the other one(designated as HIAS) was made by crossing the NIAS population with IR24; one EPI population consisted of nine single chromosome segment subtituion lines and 33 double chromosome segments subtituion lines on chromosome 8. The major objective of this study was to systematicaly analyze and confirm the genetic effects of important quantitative traits from different populations in order to identify favorable genes in indica and japonica breeding, and reveal the genetic mechanism of heterosis and epistasis in rice. Main results were given below.1. A total of 245 quantative trait loci(QTL) were detected in the RIL population, among which 222 and 178 QTL were identified in NAIS and NIAS, respectively. Among the 245 QTL, 42 were detected in four environments simultaneously, indicating their high stability across environment. Additive effects of 455 QTL were at the same direction in four environments. There were 82 marker intervals harbored more one QTL. Novel grain shape QTL were detected in four marker intervals. In genetic analysis of grain shape, four new grain shape characters were investigated by image processing system, i.e., grain area, grain circumference, grain diameter, grain roundness. Compared with manual measurement, the image analysis is much more efficient and has higher precision. Characters recorded by the image analysis provide more complete description and improve our knowledge in the genetic archeterues on grain shape.2. A total of 365 QTL were detected in populations NAIS and HAIS, and 280 QTL were detected in populations NIAS and HIAS. The identified QTL showed various degrees of dominance. From genetic effects on individual loci, we concluded that over-dominance had major role in yield heterosis, over-dominance and partial dominance were equal important in leaf shape and grain shape heterosis. In addition, heterosis was affected by environments as well.3. Among non-allelic loci, few epistasis was detected in RIL and CSSL populations. For the few epistasis identified in RIL population, the two interacting loci had non-significant additive effects. For the few epistasis identified in CSSL population, the two interacting loci had significant additive effects either. Dominance by dominance(DD) epistatic interaction were nonsignificant in hybrid lines with japonica as genetic background, nevertheless DD epistatic interaction were significant in hybrid lines with indica as genetic background. About 17% DD showed over dominance by dominance effects. In addition, epistasis was affected by environments.4. The EPI population had 9 lines of single donor segment, and 33 lines of double donor segments from choromsome 8. By using the 42 lines, a total of 292 epistatic loci were detecd, indicating that epistasis is common in rice. Though hardly detected in RIL and CSSL populations, epistatsis could be detected in well-desinged populations, such as the population of single and double chromosomal segments. Looking into epistatic loci on leaf shape and yield related traits, we found that epistatic interactions were mainly caused by non-significant QTL. When combined together, two non-significant QTL may have a large episatic effect. Epistasis exsited on grain shape, but the epistatic effect was minor, making it difficult to be detected in RIL and CSSL populations.