Application of a Novel Quantitative Trait Locus Combination to Improve Grain Shape without Yield Loss in Rice (Oryza sativa L. spp. japonica)

Abstract

Grain shape is one of the key factors deciding the yield product and the market value as appearance quality in rice (Oryza sativa L.). The grain shape of japonica cultivars in Korea is quite monotonous because the selection pressure of rice breeding programs works in consideration of consumer preference. In this study, we identified QTLs associated with grain shape to improve the variety of grain shapes in Korean cultivars. QTL analysis revealed that eight QTLs related to five tested traits were detected on chromosomes 2, 5, and 10. Among them, three QTLs-qGL2 (33.9% of PEV for grain length), qGW5 (64.42% for grain width), and qGT10 (49.2% for grain thickness)-were regarded as the main effect QTLs. Using the three QTLs, an ideal QTL combination (qGL2^P + qGW5^P + qGT10^B) could be constructed on the basis of the accumulated QTL effect without yield loss caused by the change in grain shape in the population. In addition, three promising lines with a slender grain type were selected as a breeding resource with a japonica genetic background based on the QTL combination. The application of QTLs detected in this study could improve the grain shape of japonica cultivars without any linkage drag or yield loss.

Keywords: Oryza sativa L.; QTL; grain shape; japonica-type; marker-assisted breeding; yield loss.

Figure 1

Histograms of the mean values of agronomic traits and traits related to grain shape in the population tested for 2 years. DH, days to heading; CL, culm length; PL, panicle length; PN, panicle number; GL, grain length; GW, grain width; GT, grain thickness; RLW, the ratio of length to width; TGW, 1000 grain weight. The black arrow indicates Pecos, and the white arrow indicates Boramchan.

Figure 2

The 3D scatter plots with vertical lines for grain shape, based on the following three traits: grain length (GL), grain width (GW), and grain thickness (GT) (a). Scatter plots with a correlation line for the ratio of length to width (RLW) predicted based on the GL, GW, and GT of the population. r refers to the correlation coefficient analyzed at p < 0.001 (b). Significance levels: *** p < 0.001.

Figure 3

Effect of single QTL and QTL combinations on the following four traits: GL, GW, TGW, and GT. ^a Difference between the mean value of each line with qGL2 by t-test. ** indicates significance at p < 0.01. ^b The letters (a, b, and c) showing difference within the mean values of QTL combination by Duncan multiple range test (DMRT) at a 5% significance level.

Figure 4

The phenotypic effect of the RLW and TGW by the QTL combination identified in this study. The blue bars and letters on the bar graph indicate grades involved in TWG, and red bars indicate grades involved in RLW. The red boxes of the QTL combination composed of eight groups (ty1–ty8) indicate positive QTL associated with the trait, and the gray boxes indicate negative QTL. The lower-case letters on the error bar indicate significantly different values of measured traits based on the Duncan multiple range test (DMRT) at a 5% significance level.

Figure 5

Scatter plot of the ratio of length to width (RLW) and 1000-grain weight (TGW) in 89 lines without heterozygous distribution based on the eight QTL combination types. P1: Boramchan; P2: Pecos; ty1–ty8: eight QTL combinations.

Figure 6

Grain shape of the parents, Boramchan and Pecos, were used in this study. L/W—ratio of length to width.