| Harvest index in rice is the ratio of grain yield to total biomass. Numerious studies have indicated the harvest index is closely correlated with yield. High harvest index has been considered as one of the important traits in rice breeding for super-yield. Howerve, rice breeding for high harvest index is difficult because harvest index is a quantitative trait controlled by multiple genes and easily affected by environments. It is difficult to select harvest index in rice breeding by using conventional breeding methods based on phenotyping. Understanding the genetic base of harvest index and carrying out the efficient and accurate molecular breeding are the prerequisites for efficient rice breeding for high harvest index.In the previous study, we identified four QTLs for harvest index using a recombinant inbred line (RIL) population derived from a cross between Yuexiangzhan, an indica cultivar with high harvest index and Shengbasimiao, an indica cultivar with low harvest index. These QTLs were located on chromosomes 1,3,8 and 12 and could explain 4.6~44.5% of the phenotypic variations. The positive alleles contributing to increase harvest index came from the high harvest index parent Yuexiangzhan. Among four QTLs, the QTL qHI-8 on chromosome 8 could be detected in two consecutive years, and explained 42.8% and 44.5% of the phenotypic variations, respectively. It is a stably expressed and novel major QTL, and has the important application value in rice breeding. In the present study, a NIL-F2 populations developed from a residual heterozygous line (RHL) containing a heterozygous segment surrounding the qHI-8 locus but is homozygous in other regions was used for fine mapping of qHI-8. In order to understand the relationship between harvest index and thousand grain weight, the QTL for thousand grain weight detected in this region was also fine-mapped. The results are summarized as the followings:1. An RHL carrying a heterozygous segment flanked by the SSR marker RM419 and RM477 on chromosome 8 but homozygous elsewhere in the genome was obtained by molecular analysis of RIL population. A single plant was selected and selfed to produce progenys segregating for target region in a near isogenic background. This NIL-F2 pupulation is consisted of 400 plants.2. qHI-8 was validated by QTL analysis in a NIL-F2 population consisting of 182 plants. QTL analysis defined qHI-8 to an interval of ID190-ID209 with a LOD score of 27.5. qHI-8 explained 53.9% of phenotypic variation in the NIL-F2 population. The interval of ID 190-ID209 overlapped the interval of RM502 (ID190)-RM6845 defined in primary mapping, indicating that qHI-8 was not an artifact in primary mapping using RIL population. Meantime, a QTL controlling thousand grain weight was detected with LOD score of 59.1. This QTL explained 77.6% of phenotypic variation in the NIL-F2 population, and was designated as qTGW-8.3. By using the homozygous recombinants with different genotypes and analyzing the differences in the traits of these recombinants, qHI-8 and qTGW-8 were delimited to a 140 kb interval defined by markers ID 190 and ID209.qHI-8 was validated and fine mapped in a 140kb region in this study. The results lay down a good foundation for cloning of the gene underlying qHI-8. With the validation and fine mapping of qHI-8, the molecular markers closely linked to qHI-8 will be developed. These markers can be used in rice breeding to select harvest index by genotyping instead of phenotyping. The efficiency and accuracy of selection will be improved without affected by growth period, environmental factors and so on. So, it will facilitate high-yielding breeding for varieties with high harvest index in rice through marker-assisted selection. |