利用单片段代换系研究水稻产量相关性状QTL加性及上位性效应

产量及其相关性状如单株有效穗数、千粒重、穗实粒数、穗总粒数和结实率等是水稻重要的农艺性状，了解产量及其相关性状QTL的加性及上位性效应对以分子标记聚合育种改良水稻产量具有重要意义。本文以16个单片段代换系及15个双片段代换系分析了水稻产量相关性状QTL的加性及上位性效应。共检出影响产量及其相关性状的13个QTL，包括产量性状1个、单株有效穗数1个、千粒重4个、穗实粒数4个、穗总粒数2个和结实率1个，分布于第2、第3、第4、第7和第10染色体上。此外，检出12对双基因互作。结果显示，2个正向(或负向)产量性状QTL聚合，往往会产生负向(或正向)的上位性效应，能否产生更大(或更小)的目标性状，取决于双片段遗传效应(加性效应与上位效应代数和)绝对值与单片段最大加性效应绝对值的差。本研究结果对实施高产分子标记聚合育种方法有重要参考价值。

Epistatic and Additive Effects of QTL for Yield-Related Traits Using Single Segment Substitution Lines in Rice (Oryza sativa L.)

Yield-related traits such as panicle number per plant, thousand grain weight, number of grains per panicle, number of spikelets per panicle and seed setting rate are important agronomic traits in rice. Understanding additive and epistatic effects of QTL for yield-related traits are important to increase rice yield using method of pyramiding breeding with molecular marker assisted selection. In this paper, additive and epistatic effects of QTLs for rice yield-related traits were analyzed using 16 single segment substitution lines (SSSL) and 15 double segment substitution lines (DSSL). A total of thirteen QTLs for yield-related traits were identified on the chromosomes 2, 3, 4, 7, and 10 respectively, containing one for grain yield per plant (GY), one for panicle number (PN), four for thousand grain weight (TGW), four for number of grains per panicle (NGP), two for number of spikelets per panicle (NSP) and one for seed setting rate (SSR). Furthermore, twelve pairs of digenic interactions were detected for yield-related traits. The results showed that pyramiding two QTLs with positive effects (or two QTLs with negative effects) often results in negative epistatic effects (or positive epistatic effects) in DSSL. Whether larger or smaller value of yield-related traits is produced lies on the difference between the absolute value of genetic effect (algebraic sum of additive and epistatic effects) in the DSSL and the largest value of additive effect in the SSSL. These results are important to improve yield by pyramiding favorable QTLs for yield-related traits.