Abstract: 【Objective】Accuracy mapping of the target traits is the prerequisite of the effectiveness of marker-assisted selection (MAS). 【Method】 Two previously identified quantitative trait loci (QTLs) conferring 1000-grain weight (TGW) and grain yield per plant (YP), on the same location of chromosome 6, designated tgw6.1 and yp6.1 respectively, were chosen as target QTLs to verify the QTL effects by using three segregating populations through marker-assisted selection (MAS). Three segregating populations (BC1, F2) derived from crosses between an upland rice variety (IRAT109, donor of target QTLs) with big grain weight and high yield potential, and two lowland rice varieties (recipients) with small grain weight and better grain quality, were used in this study. 【Result】The results showed that QTLs conferring bigger TGW and higher YP were genetically stabilized in different populations with different genetic backgrounds. The differences between the means of TGW of individuals carrying and not carrying donor's favorable alleles of tgw6.1 were 3.18﹣3.62 g, significant at 1% level. The phenotypic effects of MAS for tgw6.1 were 13.94%﹣18.15%. The differences between the means of YP of individuals with and without donor's favorable alleles of yp6.1 were 5.04﹣8.18 g, significant at 5% or 1% level. The phenotypic effects of MAS for yp6.1 were 34.89%﹣58.88%. Compared the phenotypic effects based on one or both of the flanking markers of target QTLs, the results showed that it was more reliable to select with both of the flanking markers when the interval of the target QTL was longer, (such as the interval of TGW and YP QTL, 13.5 cM); when the map distance between the marker and the target QTL was very short ( such as RM527 in present study, 1.5 cM ), it was also more effective to select with the marker which was nearer to the target QTL. Moreover, the phenomena of QTL with pleiotropic effects and the effectiveness of MAS for complex traits were also discussed. 【Conclusion】In this paper, the reality and accuracy mapping of TGW and YP QTLs and their phenotypic effects were verified. The feasibility of MAS based on QTL for complex traits such as drought resistance was confirmed.

Key words: lowland rice, upland rice, 1000-grain weight (TGW), grain yield per plant (YP), quantitative trait loci(QTL), verification, marker-assisted selection