水稻籼粳交DH群体收获指数及源库性状的QTL分析

以 1个水稻籼粳交 (圭 6 30 0 2 4 2 8)来源的DH群体为材料 ,利用 1张含有 2 32个标记的RFLP连锁图谱和基于混合线性模型的定位软件QTLMapper1 0对水稻收获指数及生物量、籽粒产量、库容量和株高 5个性状进行QTL分析 ,共检测到 2 1个主效应QTLs和 9对上位性互作位点。其中 ,控制籽粒产量的 3个QTLs合计贡献率为 4 2 % ,LOD值为 7 10 ;这 3个QTLs或者与收获指数的QTL同位 ,或者与生物量的QTL同位 ,且加性效应的方向一致 ,从而揭示了“籽粒产量 =生物量×收获指数”的遗传基础所在。控制收获指数的 4个QTLs合计贡献率为 4 6 % ,LOD值为 10 3;控制生物量的 4个QTLs合计贡献率为 6 4 % ,LOD值为 14 0 9;收获指数的 4个QTLs与生物量的 4个QTLs均不同位。因此 ,通过基因重组 ,可能实现控制收获指数和生物量的增效基因的聚合 ,由此获得收获指数和生物量“双高”的基因型。检测到 5个株高QTLs,其合计贡献率为 6 4 % ,LOD值为 11 6 2 ;其中 ,有 3个效应较小的QTLs与生物量、库容量和 或籽粒产量QTLs同位 ,且同位QTLs的加性效应方向一致 ;未发现株高QTLs与收获指数QTLs的同位性。由此表明 ,株高与“源 流 库”概念中的“源”和“库”在遗传上有一定程度的关联 ,而与“流”无关联。此外还发现 ,在上述同位性QTL

Characterization of QTLs for Harvest Index and Source-sink Characters in a DH Population of Rice (Oryza sativa L.)

A DH population containing 81 DH lines from an indica-japonica cross of rice and an RFLP linkage map consisting of 232 markers were used to map quantitative trait loci(QTLs) for harvest index,biomass,grain yield,sink capacity and plant height by a computer program QTLMapper1.0 based on mixed linear models. A total of 21 significant main-effect QTLs and 9 pairs of epistatic loci were detected.Of these,three detected QTLs for grain yield collectively accounted for 42% of the phenotypic variation with a LOD of 7.10.These three grain yield QTLs were corresponded either to QTLs for harvest index or QTLs for biomass in both locations and directions of additive effects,which sheds light on the genetic basis of the formula ‘grain yield=biomass×harvest index'.Four detected QTLs for harvest index collectively explained 46% of the total phenotypic variation and four QTLs for biomass jointly accounted for 64% of the trait variation.No coincidence of harvest index QTLs with any biomass QTLs was found,therefore indicating the possibility of pyramiding favorable alleles for both traits through gene recombination so as to obtain a genotype possessing both high harvest index and heavy plant biomass.Five QTLs for plant height were detected that cumulatively explained 64% of the phenotypic variation with a LOD of 11.62.Among these, three with smaller effects respectively co-located with some of the QTLs for biomass,sink capacity and/or grain yield,but not with any of harvest index QTLs,thus suggesting that plant height was to some extent directly associated with ‘source' and ‘sink' but not with ‘transportation' of the ‘source-transportation-sink' concept,at least in this genetic background and environment.In view of a somewhat low resolution of the genetic map used in the study and the fact that when plant height QTLs co-located with those for yield and/or yield related traits,these co-located QTLs were all in the same directions of additive effects,it is more likely that these QTLs co-located in a same chromosomal region might be a single QTL which have effects on multiple traits.If this is true,the above observation have led us to assume that QTLs which have pleotropic effects on yield and/or yield related traits and plant height are very different from those which had relatively large effects only on plant height.The former contribute strongly to yield and/or yield related traits but weakly to plant height while the later contribute mainly to plant height.Obviously,due to that an increase of plant height is always coupled with an increase in lodging susceptibility,discriminating between above two types of QTLs is critical in breaking the traits' undesired association in breeding for improved yield potential of rice.In addition,based on the co-location analysis of main-effect QTLs for the studied traits,five genomic regions were found to be highly associated with harvest index,biomass,sink capacity and grain yield.