摘要

To investigate the digenic epistatic effects and their environmental interactions for allelopathy in rice (Oryza saliva L.), we established a mapping population of 123 F-10 lines (recombinant inbred lines, RILs) derived from a cross between Japonica variety, Lemont (non-allelopathice rice) and Indica variety, Dular (allopathic rice). The Quantitative trait loci (QTL) analysis was done for allelopathy in the rice varieties during 2004 and 2005. Inhibition rate (IR) on the root length of the target plant bamyardgrass (Echinochloa crus-galli) was detected in (RIL) population of rice accessions by using the relay seeding co-culture bioassay method. The software QTLMapper 1.0 was applied to analyze QTLs, including QTL mapping, main genetic effect and epistatic effects as well as the QTL interaction effects with environment (QE) for allelopathy in rice. The two QTLs that had significant additive effects on the root length of target plant bamyardgrass were identified and mapped to chromosomes 2 and 5, which performed 3.69 and 3.05 LOD values and explained 6.95% and 4.35% of phenotypic variance, respectively, of which one locus performed its significant interaction with environment (AE). Epistasis analysis indicated that there were 3 pairs of QTLs in the marker intervals located on chromosome 1,3,4,5 and 10, performing significant additive x additive interactions for allelopathic effects on the root length of receiver plant bamyardgrass with positive epistatic effects, of which one pair of loci was significant in additivexadditive interactions with environment (AAE). The findings suggested that a "favorable" gene with positive additive effects might become "unfavorable (deleterious)" following transfer into a new variety due to the large negative additive x additive effects in the new genetic background. Therefore for allelopathy in rice, QTL mapping and selection experiments should place more emphasis on identifying the best multi-locus alielic combinations instead of pyramiding individual favorable QTL allefes.