Detection of QTL on panicle number in rice (Oryza sativa L.) under different densities with single segment substitution lines

Abstract

The elite indica rice variety Hua-jing-xian 74 and its 12 single segment substitution lines (SSSLs), all of which have been shown to have quantitative trait loci (QTL) for panicle number (PN), were used as crossing parents to construct a half-diallel crossing population with the aim of analyzing the expression of these QTL under different cropping densities. A total of 91 half-diallel crossing combinations were grown in two planting seasons at three cropping densities. PN was measured at the mature stage. The additive, dominant and epistatic effects of the 12 QTL as well as their interaction effects with the seasons and with the densities were estimated based on genetic effect components. Our analysis revealed that PN generally decreased with increasing cropping density. In the six single environments, eight additive QTL, nine dominant QTL and 49 pairs of epistatic QTL were detected, which were mostly associated with estimated positive, positive and negative effects, respectively. Expression of these QTL differed across planting seasons and cropping densities, implying an existing of QTL-by-environment interaction. Further analysis of the QTL effect components revealed that seven, eight and 28 pairs of QTL were present with significant additive, dominance, epistasis and/or interaction effects with densities. QTL additive and dominant effects were mostly positive, while epistatic effects were all negative. No significant QTL-by-season effects were detected. QTL Pn3-1, without any significant additive-by-density interaction effect, showed stable additive expression across densities. QTL Pn3-1, Pn3-2 and Pn6-1 showed stable dominance expression, and QTL pairs Pn2-1/Pn9, Pn2-2/Pn3-1, Pn2-2/Pn6-2, Pn3-1/Pn6-1, Pn3-1/Pn7 and Pn6-1/Pn6-3 had stable epistasis expression. The remaining significant QTL had different effects across densities. We determined that a density of 10 × 16.7 cm2 had little influence on QTL expression, that a density of 16.7 × 16.7 cm2 mostly increased QTL additive and dominant effects and decreased QTL epistatic effects and that a density of 23.3 × 16.7 cm2 had the opposite impact on QTL effects compared with 16.7 × 16.7 cm2. Additionally, the influence of density on QTL epistatic effect was generally larger than that on QTL additive or dominant effect. These results provide the information on cropping density and how it influences PN QTL expression, which may be useful information to improve rice PN via heterosis and/or QTL pyramiding.