QTL Mappling And Heterosis Prediction For Plant Height And Panicle Length In Rice

Two populations were used in the presented study, one population of 241 recombinant inbred line (RIL) was derived from female parent Zhenshan 97 and male parent Minghui 63, the other population of 240 permanent F2 was derived from the random crosses of RI. The linkage map of RIL was including 221 markers, the linkage map of PF2 was deduced from RIL. Using the mixed-model-based mapping software QTLMapper2.0, the presented study was carried out to analyze QTL with epistasis effects (including additive X additive epistasis, additive>< dominance epistasis, dominance X additive epestasis and dominance>( dominance epistasis) and QTL )( environment interaction effects for plant height and panicle length in rice. In 1998 RIL population was planted and crossed to derive PF2. The field experiments (random complete block design) were conducted in 1999 and 2000. Some conclusions of QTL mapping and heterosis prediction are as follows: 1. 14-24 QTLs were detected by full model with epistasis effects and QTL X environment for plant height and panicle length. They were all small effect genes. Many loci had no additive or dominance effects but were involved epistasis. Such loci might play the role of modifying gene that tend to active other loci or modify the action of other loci. The results that most of the QTL had both additive (or dominance) effects and epistasis main effects, some QTL had no genetic main effects but QTL X environmet interaction effects and some QTL had both genetic main effects and QTL )K environment interaction effects suggested that estimates of the QTL by the models assuming no epistasis or no QTL>< environment would result in biased estimation or QTL cannot be detected. In a specific environment, the total effect of a QTL should include the main effects plus QTL X environment interaction effects. 2. Epistasis had many features. The numbers of QTL involved in epistasis were much more than that of QTL not involved in epistasis by using RJL population or PF2 population. The numbers of QTL had additive X additive epistasis effects were more than iii additive X dominance (dominance X additive) epistasis and dominance X dominance epistasis effects. Additive X additive epistasis is the most important effects of three epistasis effects. It was common for the same locus to get involved in interactions with more than one other locus. QTL with relatively high magnitude of effects might also be involved in epistasis. The additive X additive epistasis effects of QTL played an important role in plant height and panicle length, so the additive X additive epistasis is the genetic basis of plant height and panicle length. 3. There were some differences between the results of QTL mapping for plant height by using RIL population and PF2 population except that two populations detected some QTL in the same position. The same result happened in QTL mapping for panicle length. 4. The mid-parent and better-parent heterosis in plant height ofF1, F2 and F3 were all negative. Heterosis of plant height was stable in F2 and was lower than two parents. Heterosis of hybrid was influenced by environments, so heterosis was unstable across different environments. The mid-parent and better-parent heterosis in panicle length of F1, F2 and F3 were all positive. In 1999, heterosis in panicle length was not variation among the hybrids of two parents. In 2000, there was differe...