Breeding Science
Online ISSN : 1347-3735
Print ISSN : 1344-7610
ISSN-L : 1344-7610
Research Papers
Identification of quantitative trait loci for rice grain quality and yield-related traits in two closely related Oryza sativa L. subsp. japonica cultivars grown near the northernmost limit for rice paddy cultivation
Noriko KinoshitaMasayuki KatoKei KoyasakiTakuya KawashimaTsutomu NishimuraYuji HirayamaItsuro TakamureTakashi SatoKiyoaki Kato
Author information
JOURNALS OPEN ACCESS FULL-TEXT HTML

2017 Volume 67 Issue 3 Pages 191-206

Details
Abstract

Quantitative trait loci (QTLs) associated with eating quality, grain appearance quality and yield-related traits were mapped in recombinant inbred lines (RILs) derived from closely related rice (Oryza sativa L. subsp. japonica) cultivars, Yukihikari (good eating quality) and Joiku462 (superior eating quality and high grain appearance quality). Apparent amylose content (AAC), protein content (PC), brown grain length (BGL), brown grain width (BGWI), brown grain thickness (BGT), brown grain weight per plant (BGW) and nine yield-related traits were evaluated in 133 RILs grown in four different environments in Hokkaido, near the northernmost limit for rice paddy cultivation. Using 178 molecular markers, a total of 72 QTLs were detected, including three for AAC, eight for PC, two for BGL, four for BGWI, seven for BGT, and six for BGW, on chromosomes 1, 2, 3, 4, 6, 7, 8, 9, 11 and 12. Fifteen intervals were found to harbor multiple QTLs affecting these different traits, with most of these QTL clusters located on chromosomes 4, 6, 8, 9 and 12. These QTL findings should facilitate gene isolation and breeding application for improvement of eating quality, grain appearance quality and yield of rice cultivars.

Introduction

Rice (Oryza sativa L.) grain quality has four characteristics, i.e. eating quality, appearance quality, milling quality, and nutritional quality. It is necessary for rice breeders to understand how these quality traits are inherited. The percentage of amylose in total starch, measured as apparent amylose content (AAC), is the key determinant of rice cooking properties. AAC is a complex trait in rice (Ikeno 1914) and is controlled by many genes, including Waxy (Wx) (Sano 1984), Du1 (Satoh and Omura 1981), Du2 and Du3 (Satoh and Omura 1986), and Du4 and Du5 (Yano et al. 1988). The Waxy (Wx) gene encodes granule-bound starch synthase I (GBSSI), one of the enzymes involved in amylose synthesis, and is located on rice chromosome 6 (Sano 1984). Two functional alleles, Wxa and Wxb, have been reported in rice, with Wxb mainly found in japonica cultivars, and Wxa found in indica cultivars and various wild rice species (Sano 1984, Sano 1991). Wxa and Wxb were initially identified by the amounts of their gene products (Sano 1984). The Wxa allele produces about tenfold higher levels of mRNA and protein than Wxb. As a result, AAC of japonica cultivars is almost always below 20%, whereas the AAC of indica cultivars is higher than 20%. In addition to mutant genes at the wx locus, such as Wx-mq and Wx1-1, several other QTLs for AAC have been detected (Ando et al. 2010, Sato et al. 2002). Single QTLs for AAC have each been detected on chromosomes 1, 3, 4, 5, 6 and 12 (He et al. 1999, Li et al. 2003, Septiningsih et al. 2003, Takeuchi et al. 2007, Wan et al. 2004), and two QTLs each on chromosomes 8 (Li et al. 2011, Wan et al. 2004) and 9 (Ando et al. 2010, Wan et al. 2004). In addition to AAC, PC determines eating quality; rice with high protein content is harder, less elastic and less viscous after being cooked. Fifty-five QTLs for PC have been identified on all 12 rice chromosomes (Aluko et al. 2004, Hu et al. 2004, Liu et al. 2011, Tan et al. 2001, Yoshida et al. 2002, Yu et al. 2009, Zhang et al. 2008, Zheng et al. 2011, 2012, Zhong et al. 2011).

The eating quality of rice is also influenced by environmental factors, such as air temperature during the grain filling period (Nishimura et al. 1985) and nitrogen levels in the soil (Ishima et al. 1974). In general, cool temperatures during the filling period reduce eating quality by elevating AAC (Asaoka et al. 1985). Wx gene expression and Wx protein were increased when rice plants were exposed to low temperature (18°C) (Larkin and Park 1999, Sano et al. 1985). Nitrogen level in the soil strongly affects not only yield but grain quality. Yield has been hypothesized to be related to the nitrogen supplying capacity of soil, which in turn determines grain protein content (Perez et al. 1996). Application of nitrogen fertilizer at different stages, including panicle initiation, heading, flowering, and grain filling, has been shown to strongly increase seed-storage protein content (Leesawatwong et al. 2004, 2005, Nagarajah et al. 1975, Nangju and De Datta 1970, Perez et al. 1990, 1996, Seetanun and De Datta 1973, Souza et al. 1999, Taira 1970, Vaughan et al. 1980). In contrast, application of nitrogen has also been reported to reduce AAC (Bahmaniar and Ranjbar 2007).

Grain shape, which includes gain length, width and thickness, is a key determinant of the quality of grain appearance (Huang et al. 2013), as well as being an important component of grain yield. In Japan, brown rice grains are mechanically sieved at a mesh width of 1.70–2.0 mm, depending on cultivars and locations. This sieving yields two fractions, consisting of thick (>1.7–2.0 mm) and thin (<1.7–2.0 mm) grains, with the thicker grains generally marketed. More recently, 2.0 mm mesh is increasingly used to separate out thin brown rice grains, making brown rice of thickness >2.0 mm essential for rice cultivars in Japan. Grain shape is also widely accepted as a complex trait controlled by multiple genes, each with small effects. Hulls cover rice grains. Brown grain length (BGL) and brown grain width (BGWI) are fixed as long as the panicle is normally differentiated. Thus, BGL and BGWI are mainly controlled by genotype. However, brown grain thickness (BGT) is thought to be largely affected by filling degree, which is considerably affected by the environment (Bai et al. 2010). Extensive efforts to determine the genetic basis of grain shape have used forward and reverse genetic strategies. Initial studies focused on characterizing mutants and the expression of major genes associated with grain size. These include, for example, the genes Lk-f, which is associated with long kernel size (Takeda and Saito 1980), and Mi, which is associated with short kernel size (Takeda and Saito 1977). Alternatively, quantitative trait locus (QTL) analysis based on genome wide mapping has been widely used over the past 20 years to map genes associated with rice grain shape. To date, nearly 200 QTLs for grain length and grain width have been reported (reviewed by Hunang et al. 2013). More recently, Nagata et al. (2015) reported a total of 130 QTLs for grain length and grain width using a single chromosome segment substitution line population and advance backcross populations. However, understanding of the genetic control of BGT remains very limited. In addition, grain shape is a key determinant of grain yield. In general, a drastic increase in grain size usually does not increase grain productivity proportionally, owing to reductions in both grain filling and grain quality resulting from imbalances between sink and source potentials (Peng et al. 2008, Takai et al. 2013, Takita 1983). Therefore, grain shape should be improved by using appropriate QTL alleles to maintain an appropriate balance between sinks and sources, thus allowing an increase in grain yield. Rice yield traits are complex and governed by multiple QTLs (reviewed by Miura et al. 2011). Most QTLs for yield traits show small genetic effects and are difficult to identify. These minor QTLs play a vital role in regulating yield traits and are widely utilized in commercial rice varieties, making identification of these QTLs beneficial for breeding. Dissecting the genetic basis of yield related traits by QTL mapping could facilitate the breeding of high yield varieties.

Japan has a long history of breeding temperate japonica rice for growth during the summer monsoon season at higher latitudes. The rice cultivars grown in Hokkaido (45–42°N), the northernmost region of rice paddy cultivation in Japan and one of the northernmost limits of rice cultivation in the world, have a relatively short alternative breeding history. Following improvements in rice production, such that Japan’s rice self-sufficiency approached 100%, the main breeding objective was changed from high yield to good eating quality (Horie et al. 2005). However, the environmental conditions in Hokkaido, low temperature and high nitrogen level, are not suitable for the production of rice with good eating quality for Japanese consumers (Inatsu 1988). Nonetheless, intensive selection pressures in Hokkaido rice breeding programs over the last three decades have focused on improving the eating quality of cooked rice. This has resulted in the stable production of rice with good eating and grain appearance qualities (Kinoshita 2013). The first good eating quality rice cultivar in Hokkaido, Yukihikari, released in 1981, was derived from the progeny of crosses between Hokkaido landraces. The eating quality of Yukihikari was further improved by inclusion of the elite Japanese cultivar Koshihikari, released for cultivation on Honshu, the main island of Japan, and other good eating quality cultivars. One recent breeding line, Joiku462, derived from the progeny of Yukihikari and released in 2009, has shown superior eating and grain appearance qualities. Less is known, however, about the QTLs associated with improvements in eating quality, grain appearance quality and yield potential in Hokkaido rice cultivars grown under regional environmental conditions. In the present study, QTLs for traits related to AAC, PC, grain shape and grain yield were mapped in a population of recombinant inbred lines (RILs) of a cross between the two closely related cultivars, Yukihikari and Joiku462, using our previously developed PCR-based markers from InDel polymorphisms and single nucleotide polymorphisms (SNPs) (Kinoshita et al. 2016, Takano et al. 2014b).

Materials and Methods

Plant materials

Oryza sativa L. ssp. japonica cv. Yukihikari and Joiku462 were used as parental lines. Both were grown in Hokkaido, Japan, with Joiku462, released in 2009, being a progeny of Yukihikari, released in 1981. The 133 RILs (F10 and F11) were developed by the single seed descent (SSD) method of progenies derived from a cross between Yukihikari and Joiku462 (Kinoshita et al. 2016). The F10 and F11 RIL populations were used for field trials in 2014 and 2015, respectively.

Trait measurements

Days to heading (DTH) were defined as the number of days from sowing to more than 50% of plants with heading, based on visual observation. At maturity, panicle length (PL), culm length (CL) and panicle number (PN) of five or more randomly chosen plants of each parental line or RIL were measured and averaged. Grain number per plant (GN), filled grain number per plant (FGN), grain number per panicle (GNP) and unfilled grain ratio (UFG) of two or more randomly chosen plants of each parental line or RIL were measured and averaged. To measure brown grain weight (BGW), the grains of more than eight plants of each parental line or RIL were pooled, air-dried to a moisture content of 15–16%, and weighed, and the average number of grains per plant was calculated. The combined weight of two samples of 500 randomly chosen brown rice grains per line was defined as the 1000 brown grain weight (TBGW). Brown grain length (BGL), brown grain width (BGWI) and brown grain thickness (BGT) were measured in 500 randomly chosen brown rice grains from each line using a Satake Grain Scanner (RGQI10B, Satake, Hiroshima, Japan) and averaged. More than 50 grams of brown rice were polished to a yield of ~90% in a rice mill (SKM5B(1); Satake, Hiroshima, Japan). The apparent amylose content (AAC) of polished rice from each line was evaluated as described (Juliano et al. 1965), and duplicated protein contents (PC) of polished rice of each line were determined using an Infratec™ 1241 Grain Analyzer (Foss, Hillerød, Denmark).

Information on field experiments and QTL analyses are presented in Supplemental Text 1.

Results

Trait performance of parents and RIL populations

Table 1 shows the phenotypic variations of parental lines and the RIL population for 15 traits across four environments. Eight traits, DTH, AAC, PC, TBGW, BGL, BGWI, BGT and GNP, differed significantly in the two parental lines in three or more environments (P < 0.05 each). Joiku462 headed 2.4 to 5 d earlier than Yukihikari in 2014 in Pippu (2014P) and in 2014 and 2015 in Sapporo (2014S and 2015S, respectively). AAC and PC were lower, and TBGW, BGL, BGWI and BGT were higher, in Joiku462 than in Yukihikari across all four environments. In contrast, GNP of Joiku462 was lower than that of Yukihikari in 2014P, 2014S and 2015S. In addition, five traits, GN, FGN, UFG and PL, were significantly lower, and two traits, PN and CL, were significantly higher in Joiku462 than in Yukihikari in one or two environments (P < 0.05 each). Taken together, these findings indicate that Joiku462 showed improvements in eating quality, with lower amylose and protein contents and better grain appearance, along with larger grains and early heading. Although Joiku462 tended to have an increased number of panicles, its panicles were shorter, with fewer grains per panicle, than Yukihikari.

Table 1 Phenotypic data for eating quality, grain appearance quality and yield related traits of the 133 RILs and parents, Yukihikari and Joiku462 in 2014P, 2014S, 2015P and 2015S
Trait Trait description Environment Parental mean RILs
Yukihikari Joiku462 Difference (J-Y) Mean Min Max
DTH Day to heading 2014P 86.3 82.0 −4.3** 87.0 74.0 98.0
2014S 83.7 81.3 −2.4* 84.0 69.0 96.0
2015P 109.3 106.7 −2.7 108.6 99.0 116.0
2015S 92.0 87.0 −5.0** 99.4 87.0 108.0
AAC Apparent amylose content (%) 2014P 19.82 17.73 −2.09*** 18.35 13.53 22.51
2014S 19.27 17.45 −1.82** 18.21 13.16 23.01
2015P 22.57 19.94 −2.63*** 21.07 16.30 25.20
2015S 21.12 19.27 −1.85*** 19.10 13.90 22.80
PC Protein content (%) 2014P 6.50 5.77 −0.73** 6.10 5.20 9.10
2014S 7.27 6.27 −1.00* 7.00 5.40 8.90
2015P 5.95 5.59 −0.36* 5.93 4.80 8.00
2015S 7.42 6.95 −0.47* 6.48 5.20 8.40
BGW Brown grain weight per plant (g) 2014P 32.8 33.4 0.6 32.1 12.9 43.5
2014S 23.6 28.2 4.6 25.5 10.6 41.2
2015P 31.0 28.5 −2.6 26.4 14.5 39.3
2015S 31.6 38.0 6.4 25.4 11.4 47.4
TBGW Thousand brown grain weight (g) 2014P 22.6 25.0 2.4*** 23.1 19.4 26.9
2014S 22.6 24.9 2.3** 23.4 19.8 26.8
2015P 22.1 24.7 2.7*** 22.5 19.2 25.6
2015S 22.1 25.2 3.0*** 22.3 18.0 26.7
BGL Brown grain length (mm) 2014P 4.99 5.28 0.30*** 5.15 4.81 5.55
2014S 4.95 5.20 0.25*** 5.12 4.81 5.52
2015P 4.94 5.26 0.32*** 5.09 4.70 5.43
2015S 4.95 5.26 0.31*** 5.07 4.65 5.39
BGWI Brown grain width (mm) 2014P 2.97 3.02 0.05** 2.94 2.71 3.19
2014S 2.94 3.01 0.06** 2.95 2.66 3.18
2015P 2.96 3.02 0.06*** 2.93 2.70 3.19
2015S 2.99 3.04 0.05* 2.97 2.76 3.21
BGT Brown grain thickness (mm) 2014P 1.99 2.06 0.07*** 2.00 1.90 2.09
2014S 1.98 2.04 0.06** 2.00 1.90 2.09
2015P 2.00 2.07 0.06*** 2.01 1.93 2.10
2015S 2.02 2.08 0.06* 2.01 1.92 2.09
GN Grain number per plant 2014P 2602.0 1924.0 −678.0** 2200.0 1085.0 3168.0
2014S 1260.3 1151.3 −109.0 1249.0 607.0 2724.0
2015P 1178.0 1077.2 −100.8 1484.9 537.0 3034.0
2015S 1795.5 1799.7 4.2 1513.7 677.0 2504.0
GNP Grain number per panicle 2014P 84.4 51.7 −32.7** 70.1 34.1 100.3
2014S 75.0 51.9 −23.1** 65.9 27.1 100.9
2015P 72.9 55.1 −17.8 60.2 30.0 99.0
2015S 83.2 60.1 −23.1* 65.3 35.0 96.0
FGN Filled grain numer per plant 2014P 2299.0 1818.3 −480.7** 2028.0 1048.0 2814.0
2014S 1054.7 1082.0 27.3 1123.0 518.0 2510.0
2015P 1370.8 1245.3 −125.5 1411.9 524.0 2884.0
2015S 1714.3 1735.9 21.6 1400.2 645.0 2303.0
UFG Unfilled grain ratio (%) 2014P 11.5 5.4 −6.1 7.3 1.9 22.3
2014S 14.1 5.7 −8.5 8.0 2.4 29.6
2015P 5.6 3.2 −2.4** 5.0 1.1 15.7
2015S 4.6 3.7 −0.9 5.7 2.2 29.9
PL Panicle length (cm) 2014P 19.92 18.06 −1.87* 18.20 14.50 22.02
2014S 18.80 17.84 −0.96 17.75 11.68 21.45
2015P 19.32 17.81 −1.51* 17.36 13.80 22.30
2015S 18.72 19.64 0.92 17.75 13.80 20.70
PN Panicle number 2014P 26.9 30.8 3.9** 27.2 20.3 40.2
2014S 19.5 23.8 4.3 19.8 12.3 29.0
2015P 21.3 25.2 3.9*** 23.5 16.0 38.0
2015S 21.0 29.2 8.2 22.3 12.0 38.0
CL Culm length (cm) 2014P 74.11 76.22 2.11 75.57 51.85 90.90
2014S 67.05 68.39 1.34 63.45 40.15 83.58
2015P 71.89 74.75 2.86 72.75 49.20 90.10
2015S 69.34 76.10 6.76* 70.52 50.00 83.90
*, ** and ***  mean the significance levels 5%, 1% and 0.1% between Yukihikari and Joiku462 respectively.

2014P and 2014S are Pippu and Sapporo in 2014 and 2015P and 2015S are Pippu and Sappro in 2015.

AAC of the RIL population segregated as a bimodal distribution across the four environments (Fig. 1). The remaining fourteen traits of the RIL population segregated continuously across the four environments. Transgressive segregants, with higher or lower values than their respective parents, were observed for all traits across the four environments (Table 1, Fig. 1).

Fig. 1

Frequency distribution of days to heading (DTH), apparent amylose content (AAC), protein content (PC), brown grain weight per plant (BGW), 1000 brown grain weight (TBGW), brown grain length (BGL), brown grain width (BGWI), brown grain thickness (BGT), grain number per plant (GN), grain number per panicle (GNP), filled grain number per plant (FGN), unfilled grain ratio (UFG), panicle length (PL), panicle number (PN) and culm length (CL) in the RILs derived from the cross between Yukihikari and Joiku462 grown in Pippu and Sapporo in 2014 and 2015. White and black arrows indicate the mean values for Yukihikari and Joiku462, respectively.

Fig. 1

Frequency distribution of days to heading (DTH), apparent amylose content (AAC), protein content (PC), brown grain weight per plant (BGW), 1000 brown grain weight (TBGW), brown grain length (BGL), brown grain width (BGWI), brown grain thickness (BGT), grain number per plant (GN), grain number per panicle (GNP), filled grain number per plant (FGN), unfilled grain ratio (UFG), panicle length (PL), panicle number (PN) and culm length (CL) in the RILs derived from the cross between Yukihikari and Joiku462 grown in Pippu and Sapporo in 2014 and 2015. White and black arrows indicate the mean values for Yukihikari and Joiku462, respectively.

Fig. 1

Frequency distribution of days to heading (DTH), apparent amylose content (AAC), protein content (PC), brown grain weight per plant (BGW), 1000 brown grain weight (TBGW), brown grain length (BGL), brown grain width (BGWI), brown grain thickness (BGT), grain number per plant (GN), grain number per panicle (GNP), filled grain number per plant (FGN), unfilled grain ratio (UFG), panicle length (PL), panicle number (PN) and culm length (CL) in the RILs derived from the cross between Yukihikari and Joiku462 grown in Pippu and Sapporo in 2014 and 2015. White and black arrows indicate the mean values for Yukihikari and Joiku462, respectively.

The correlation coefficients of each pairwise combination are summarized in Supplemental Table 1 and presented in Supplemental Text 2.

QTL identification

Significant QTLs were identified for all 15 traits associated with eating quality, grain appearance quality and yield related traits (Table 2). A total of 72 QTLs were identified, including five for DTH, three for AAC, eight for PC, six for BGW, seven for TBGW, two for BGL, four for BGWI, seven for BGT, five for GN, five for GNP, six for FGN, one for UFG, six for PL, three for PN and four for CL. These QTLs were distributed on 10 chromosomes, including chromosomes 1, 2, 3, 4, 6, 7, 8, 9, 11 and 12, with the majority clustered on chromosomes 4, 6, 8, 9 and 12 (Fig. 2). The phenotypic variations attributed to each QTL ranged from 2.3–75.6%. Twelve of these QTLs (16.7%), qDTH6, qDTH8, qAAC4, qAAC8, qAAC9, qTBGW4, qBGL4, qBGL11, qBGWI9, qBGT1, qGNP6 and qGNP12, were consistently identified across all four environments. Seven QTLs (9.7%), qPC8, qTBGW9, qTBGW11, qBGWI12.2, qBGT4.2, qPL8 and qCL12, were identified in three environments, and 15 QTLs (20.8%), qDTH3, qDTH12, qPC6.2, qBGW6.1, qBGW6.2, qBGW8, qBGW12, qTBGW1, qTBGW8, qBGT12, qUFG6, qPL6.1, qPL6.2, qCL6 and qCL8, in two environments. These 15 QTLs were classified into four groups. Six of these QTLs, qPC6.2, qBGW6.1, qBGW6.2, qTBGW8, qPL6.2 and qCL6, were specific to Pippu (2014P and 2015P); four, qDTH3, qDTH12, qBGW8 and qBGW12, were specific to Sapporo (2014S and 2015S); two, qBGT12 and qCL8, were specific to 2014 (2014P and 2014S); and three, qTBGW1, qUFG6 and qPL6.1, were specific to 2015 (2015P and 2015S). The remaining 38 QTLs (52.8%), qDTH7, qPC1, qPC2, qPC3, qPC6.1, qPC12.1, qPC12.2, qBGW1, qBGW9, qTBGW12.1, qTBGW12.2, qBGWI8, qBGWI12.1, qBGT4.1, qBGT7, qBGT8, qBGT9, qGN2, qGN3, qGN6, qGN8, qGN9, qGNP4, qGNP7, qGNP8, qFGN1, qFGN2, qFGN3, qFGN6, qFGN8, qFGN9, qPL4, qPL11, qPL12, qPN7, qPN9, qPN12 and qCL7, were specific to one environment each.

Table 2 Putative QTLs for eating quality, grain appearance quality and yield related traits detected in the RIL population derived from the cross between Yukihikari and Joiku462 grown in Pippu and Sapporo in 2014 and 2015
Trait QTL Chr. Environment LOD threshold Nearest marker Marker interval LOD PVE(%) Additive effect Donor of positive allele
Marker Physical position (Mb)a Marker Physical position (Mb)a
DTH qDTH3 3 2014P 2.9 YJInDel-129 24.5 YJInDel-129–YJInDel-130 24.5–24.7 3.54 5.2 1.29 Joiku462
2015S 2.9 YJInDel-129 24.5 YJInDel-129–YJInDel-130 24.5–24.7 3.50 5.4 1.69 Joiku462
qDTH6 6 2014P 2.9 YJInDel-218 9.7 YJInDel-208–YJInDel-218 7.9–9.7 19.23 33.4 3.51 Yukihikari
2014S 2.9 YJInDel-208 7.9 YJInDel-208–YJInDel-218 7.9–9.7 9.36 21.3 2.55 Yukihikari
2015P 3.0 YJInDel-218 9.7 YJInDel-208–YJInDel-218 7.9–9.7 19.60 36.5 2.65 Yukihikari
2015S 2.9 YJInDel-218 9.7 YJInDel-208–YJInDel-218 7.9–9.7 21.69 43.3 4.90 Yukihikari
qDTH7 7 2014P 2.9 YJInDel-301 27.0 YJInDel-301–YJInDel-304 27.0–29.0 3.79 5.4 1.42 Yukihikari
qDTH8 8 2014P 2.9 YJInDel-315 3.7 YJInDel-306–YJInDel-320 3.0–4.1 7.88 12.5 1.92 Joiku462
2014S 2.9 YJInDel-321 5.1 YJInDel-320–YJInDel-321 4.1–5.1 10.53 25.5 2.77 Joiku462
2015P 3.0 YJInDel-320 4.1 YJInDel-320–YJInDel-321 4.1–5.1 9.76 15.8 1.71 Joiku462
2015S 2.9 YJInDel-321 5.1 YJInDel-320–YJInDel-321 4.1–5.1 5.83 9.0 2.23 Joiku462
qDTH12 12 2014P 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 4.05 6.0 1.37 Yukihikari
2015S 2.9 YJInDel-515 25.4 YJInDel-515 25.4 3.51 5.4 1.60 Yukihikari
AAC qAAC4 4 2014P 3.0 YJInDel-171 32.8 YJInDel-170–YJInDel-175 32.7–33.1 3.96 2.3 0.33 Joiku462
2014S 3.0 YJInDel-171 32.8 YJInDel-170–YJInDel-175 32.7–33.1 5.22 3.9 0.46 Joiku462
2015P 3.0 YJInDel-171 32.8 YJInDel-170–YJInDel-171 32.7–32.8 4.89 3.3 0.42 Joiku462
2015S 2.9 YJInDel-171 32.8 YJInDel-170–YJInDel-171 32.7–32.8 4.54 3.1 0.42 Joiku462
qAAC8 8 2014P 3.0 YJInDel-320 4.1 YJInDel-320–YJInDel-321 4.1–5.1 6.19 3.5 0.45 Joiku462
2014S 3.0 YJInDel-320 4.1 YJInDel-320–YJInDel-321 4.1–5.1 5.72 4.2 0.53 Joiku462
2015P 3.0 YJInDel-315 3.7 YJInDel-306–YJInDel-315 3.0–3.7 6.61 4.5 0.50 Joiku462
2015S 2.9 YJInDel-315 3.7 YJInDel-306–YJInDel-315 3.0–3.7 5.50 3.9 0.48 Joiku462
qAAC9 9 2014P 3.0 YJInDel-351 3.7 YJInDel-351 3.7 51.20 75.6 1.90 Yukihikari
2014S 3.0 YJInDel-351 3.7 YJInDel-351 3.7 43.72 71.6 1.96 Yukihikari
2015P 3.0 YJInDel-351 3.7 YJInDel-351 3.7 46.84 71.9 1.92 Yukihikari
2015S 2.9 YJInDel-351 3.7 YJInDel-351 3.7 45.99 73.3 1.99 Yukihikari
PC qPC1 1 2015S 2.9 YJInDel-536_2 29.4 YJInDel-34–YJInDel-536_2 26.2–29.4 4.03 12.1 0.26 Joiku462
qPC2 2 2014P 2.8 YJInDel-65 9.9 YJInDel-61–YJInDel-67 8.0–13.5 3.63 6.5 0.18 Yukihikari
qPC3 3 2014P 2.8 YJInDel-129 24.5 YJInDel-128–YJInDel-130 22.8–24.7 2.99 5.3 0.16 Yukihikari
qPC6.1 6 2014P 2.8 YJInDel-197 0.8 YJInDel-197–YJInDel-206 0.8–2.0 4.28 7.7 0.20 Yukihikari
qPC6.2 6 2014P 2.8 YJInDel-208 7.9 YJInDel-207–YJInDel-208 5.2–7.9 11.13 22.3 0.42 Joiku462
2015P 2.9 YJInDel-208 7.9 YJInDel-208–YJInDel-218 7.9–9.7 8.65 24.4 0.32 Joiku462
qPC8 8 2014P 2.8 YJInDel-315 3.7 YJInDel-306–YJInDel-320 3.0–4.1 5.71 10.5 0.23 Yukihikari
2014S 2.9 YJInDel-324 5.7 YJInDel-321–YJInDel-324 5.1–5.7 4.03 13.3 0.29 Yukihikari
2015P 2.9 YJInDel-324 5.7 YJInDel-324–YJInDel-340 5.7–8.6 3.28 8.5 0.20 Yukihikari
qPC12.1 12 2015S 2.9 YJInDel-1529 12.4 YJInDel-1529 12.4 3.20 9.7 0.19 Yukihikari
qPC12.2 12 2014P 2.8 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 3.86 6.9 0.19 Joiku462
BGW qBGW1 1 2014P 2.9 YJInDel-536_2 29.4 YJInDel-34–YJInDel-35 26.2–33.9 4.36 11.3 1.72 Yukihikari
qBGW6.1 6 2014P 2.9 YJInDel-197 0.8 YJInDel-197–YJInDel-206 0.8–2.0 5.97 16.0 2.19 Joiku462
2015P 2.9 YJInDel-197 0.8 YJInDel-197–YJInDel-206 0.8–2.0 5.43 10.4 1.65 Joiku462
qBGW6.2 6 2014P 2.9 YJInDel-208 7.9 YJInDel-207–YJInDel-208 5.2–7.9 3.81 9.7 2.06 Yukihikari
2015P 2.9 YJInDel-208 7.9 YJInDel-207–YJInDel208 5.2–7.9 9.15 18.3 2.63 Yukihikari
qBGW8 8 2014S 2.9 YJInDel-321 5.1 YJInDel-320–YJInDel-321 4.1–5.1 2.96 8.6 1.82 Joiku462
2015P 2.9 YJInDel-315 3.7 YJInDel-315 3.7 6.22 12.1 1.68 Joiku462
qBGW9 9 2015P 2.9 YJInDel-356 5.7 YJInDel-356–YJInDel358 5.7–9.4 3.68 6.0 1.35 Yukihikari
qBGW12 12 2014S 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 3.92 11.6 2.02 Yukihikari
2015P 2.9 YJInDel-515 25.4 YJInDel-515 25.4 4.21 7.9 1.39 Yukihikari
TBGW qTBGW1 1 2015P 2.9 YJInDel-9 7.1 YJInDel-9–YJInDel-12 7.1–8.7 3.67 6.9 0.38 Joiku462
2015S 2.9 YJInDel-9 7.1 YJInDel-9–YJInDel-12 7.1–8.7 3.45 9.7 0.54 Joiku462
qTBGW4 4 2014P 3.0 YJInDel-610 28.2 YJInDel-161–YJInDel-610 28.0–28.2 6.45 13.0 0.54 Joiku462
2014S 2.8 YJInDel-610 28.2 YJInDel-610–YJInDel-162 28.2–29.2 6.00 15.3 0.57 Joiku462
2015P 2.9 YJInDel-610 28.2 YJInDel-610–YJInDel-162 28.2–29.2 6.68 13.1 0.52 Joiku462
2015S 2.9 YJInDel-610 28.2 YJInDel-610–YJInDel-162 28.2–29.2 5.50 15.8 0.69 Joiku462
qTBGW8 8 2014P 3.0 YJInDel-347 18.4 YJInDel-347–YJInDel-724_2 18.4–21.8 4.73 9.1 0.61 Joiku462
2015P 2.9 YJInDel-347 18.4 YJInDel-347–YJInDel-724_2 18.4–21.8 3.18 5.3 0.46 Joiku462
qTBGW9 9 2014P 3.0 YJInDel-355 5.7 YJInDel-353–YJInDel-355 5.4–5.7 5.45 10.9 0.49 Joiku462
2014S 2.8 YJInDel-351 3.7 YJInDel-351 3.7 3.58 8.8 0.42 Joiku462
2015P 2.9 YJInDel-351 3.7 YJInDel-351 3.7 4.12 7.8 0.38 Joiku462
qTBGW11 11 2014P 3.0 YJInDel-442 19.8 YJInDel-441–YJInDel-442 18.8–19.8 6.40 12.9 0.55 Yukihikari
2014S 2.8 YJInDel-442 19.8 YJInDel-441–YJInDel-442 18.8–19.8 3.93 9.6 0.47 Yukihikari
2015P 2.9 YJInDel-442 19.8 YJInDel-441–YJInDel-442 18.8–19.8 3.53 6.6 0.37 Yukihikari
qTBGW12.1 12 2014P 3.0 YJInDel-504 21.6 YJInDel-502–YJInDel-504 19.3–21.6 3.10 5.9 0.41 Yukihikari
qTBGW12.2 12 2015P 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 4.59 8.7 0.45 Yukihikari
BGL qBGL4 4 2014P 2.9 YJInDel-610 28.2 YJInDel-161–YJInDel-610 28.0–28.2 9.80 22.3 0.07 Joiku462
2014S 3.0 YJInDel-161 28.0 YJInDel-161–YJInDel-610 28.0–28.2 7.50 17.9 0.07 Joiku462
2015P 3.0 YJInDel-161 28.0 YJInDel-161–YJInDel-610 28.0–28.2 8.63 18.3 0.06 Joiku462
2015S 3.0 YJInDel-610 28.2 YJInDel-161–YJInDel-610 28.0–28.2 8.14 18.8 0.06 Joiku462
qBGL11 11 2014P 2.9 YJInDel-442 19.8 YJInDel-441–YJInDel-442 18.8–19.8 9.00 20.4 0.07 Yukihikari
2014S 3.0 YJInDel-442 19.8 YJInDel-441–YJInDel-442 18.8–19.8 8.50 21.0 0.07 Yukihikari
2015P 3.0 YJInDel-442 19.8 YJInDel-441–YJInDel-442 18.8–19.8 7.76 16.3 0.06 Yukihikari
2015S 3.0 YJInDel-442 19.8 YJInDel-441–YJInDel-442 18.8–19.8 9.69 22.9 0.07 Yukihikari
BGWI qBGWI8 8 2015P 3.0 YJInDel-724_2 21.8 YJInDel-347–YJInDel-724_2 18.4–21.8 3.05 7.5 0.03 Joiku462
qBGWI9 9 2014P 3.0 YJInDel-351 3.7 YJInDel-351 3.7 3.80 10.9 0.03 Joiku462
2014S 2.9 YJInDel-351 3.7 YJInDel-351 3.7 2.90 8.4 0.03 Joiku462
2015P 3.0 YJInDel-351 3.7 YJInDel-351 3.7 4.80 12.5 0.03 Joiku462
2015S 2.9 YJInDel-351 3.7 YJInDel-351 3.7 3.38 9.8 0.03 Joiku462
qBGWI12.1 12 2014P 3.0 YJInDel-504 21.6 YJInDel-504–YJInDel-510 21.6–23.6 3.60 10.1 0.03 Yukihikari
qBGWI12.2 12 2014S 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 3.90 11.5 0.03 Yukihikari
2015P 3.0 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 3.79 9.7 0.03 Yukihikari
2015S 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 4.02 11.7 0.03 Yukihikari
BGT qBGT1 1 2014P 2.9 YJInDel-896 42.2 YJSNP-4641–YJInDel-896 39.3–42.2 2.90 6.4 0.01 Joiku462
2014S 2.8 YJInDel-896 42.2 YJInDel-896–YJInDel-47 42.2–44.8 2.80 7.0 0.01 Joiku462
2015P 3.0 YJInDel-896 42.2 YJInDel-896–YJInDel-47 42.2–44.8 4.83 12.6 0.01 Joiku462
2015S 2.9 YJInDel-896 42.2 YJInDel-896–YJInDel-47 42.2–44.8 4.11 12.0 0.02 Joiku462
qBGT4.1 4 2014S 2.8 YJInDel-158 23.9 YJInDel-158–YJInDel-160 23.9–27.9 3.80 9.7 0.01 Yukihikari
qBGT4.2 4 2014S 2.8 YJInDel-162 29.2 YJInDel-610–YJInDel-162 28.2–29.2 4.20 11.0 0.01 Joiku462
2015S 2.9 YJInDel-162 29.2 YJInDel-162 29.2 4.37 13.1 0.01 Joiku462
2015P 3.0 YJInDel-162 29.2 YJInDel-610–YJInDel-162 28.2–29.2 5.42 14.6 0.01 Joiku462
qBGT7 7 2014P 2.9 YJInDel-276 8.7 YJInDel-273–YJInDel-276 6.9–8.7 3.60 7.9 0.01 Joiku462
qBGT8 8 2014P 2.9 YJInDel-340 8.6 YJInDel-340–YJInDel-341 8.6–17.0 3.60 7.9 0.01 Joiku462
qBGT9 9 2014P 2.9 YJInDel-351 3.7 YJInDel-351 3.7 5.50 12.7 0.01 Joiku462
qBGT12 12 2014P 2.9 YJInDel-504 21.6 YJInDel-502–YJInDel-504 19.3–21.6 3.60 7.9 0.01 Yukihikari
2014S 2.8 YJInDel-504 21.6 YJInDel-502–YJInDel-510 19.3–23.6 3.30 8.5 0.01 Yukihikari
GN qGN2 2 2014S 3.0 YJInDel-65 9.9 YJInDel-65–YJInDel-67 9.9–13.5 3.24 8.5 127.58 Joiku462
qGN3 3 2014S 3.0 YJInDel-127 22.0 YJInDel-126–YJInDel-128 16.3–22.8 3.70 9.9 118.43 Joiku462
qGN6 6 2014P 2.9 YJInDel-208 7.9 YJInDel-207–YJInDel-218 5.2–9.7 4.11 11.8 126.81 Yukihikari
qGN8 8 2014S 3.0 YJInDel-315 3.7 YJInDel-306–YJInDel-315 3.0–3.7 4.34 11.7 138.45 Joiku462
qGN9 9 2014P 2.9 YJInDel-355 5.7 YJInDel-353–YJInDel-355 5.4–5.7 4.21 12.1 130.13 Yukihikari
GNP qGNP4 4 2014P 2.9 YJInDel-165 29.8 YJInDel-165–YJInDel-167 29.8–31.9 3.12 6.9 3.35 Yukihikari
qGNP6 6 2014P 2.9 YJInDel-208 7.9 YJInDel-208–YJInDel-218 7.9–9.7 7.60 17.6 5.68 Yukihikari
2014S 3.1 YJInDel-218 9.7 YJInDel-208–YJInDel-218 7.9–9.7 4.09 10.1 4.81 Yukihikari
2015P 2.9 YJInDel-218 9.7 YJInDel-208–YJInDel-218 7.9–9.7 3.10 9.0 4.06 Yukihikari
2015S 2.9 YJInDel-218 9.7 YJInDel-208–YJInDel-218 7.9–9.7 5.46 16.2 5.17 Yukihikari
qGNP7 7 2014P 2.9 YJInDel-304 29.0 YJInDel-301–YJInDel-304 27.0–29.0 4.27 9.7 3.91 Yukihikari
qGNP8 8 2014S 3.1 YJInDel-321 5.1 YJInDel-320–YJInDel-321 4.1–5.1 4.82 11.9 5.27 Joiku462
qGNP12 12 2014P 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 3.82 8.5 4.14 Yukihikari
2014S 3.1 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 4.79 12.0 5.19 Yukihikari
2015P 2.9 YJInDel-515 25.4 YJInDel-515 25.4 3.92 11.5 4.49 Yukihikari
2015S 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 3.90 11.3 4.42 Yukihikari
FGN qFGN1 1 2014P 3.0 YJInDel-12 8.7 YJInDel-9–YJInDel-16 7.1–9.9 3.26 8.4 100.86 Yukihikari
qFGN2 2 2014S 2.9 YJInDel-67 13.5 YJInDel-65–YJInDel-67 9.9–13.5 3.49 9.2 115.75 Joiku462
qFGN3 3 2014S 2.9 YJInDel-127 22.0 YJInDel-126–YJInDel-128 16.3–22.8 4.54 12.4 116.82 Joiku462
qFGN6 6 2014P 3.0 YJInDel-208 7.9 YJInDel-208–YJInDel-218 7.9–9.7 3.45 8.9 107.24 Yukihikari
qFGN8 8 2014S 2.9 YJInDel-315 3.7 YJInDel-306–YJInDel-315 3.0–3.7 4.07 11.0 113.99 Joiku462
qFGN9 9 2014P 3.0 YJInDel-351 3.7 YJInDel-351 3.7 5.16 13.7 129.63 Yukihikari
UFG qUFG6 6 2015P 2.9 YJInDel-218 9.7 YJInDel-218–YJInDel-230 9.7–11.7 3.92 12.7 1.21 Joiku462
2015S 2.3 YJInDel-208 7.9 YJInDel-208–YJInDel-218 7.9–9.7 2.83 9.8 1.18 Joiku462
PL qPL4 4 2015P 3.1 YJInDel-161 28.0 YJInDel-161–YJInDel-610 28.0–28.2 3.29 7.1 0.48 Joiku462
qPL6.1 6 2015P 3.1 YJInDel-208 7.9 YJInDel-207–YJInDel-208 5.2–7.9 7.06 16.2 0.84 Yukihikari
2015S 3.0 YJInDel-208 7.9 YJInDel-208–YJInDel-218 7.9–9.7 5.98 17.3 0.68 Yukihikari
qPL6.2 6 2014P 3.0 YJInDel-255 23.6 YJInDel-244–YJInDel-255 22.1–23.6 4.50 13.2 0.56 Yukihikari
2015P 3.1 YJInDel-255 23.6 YJInDel-244–YJInDel-255 22.1–23.6 3.77 8.2 0.54 Yukihikari
qPL8 8 2014S 3.0 YJInDel-321 5.1 YJInDel-321–YJInDel-324 5.1–5.7 5.40 15.0 0.81 Joiku462
2015P 3.1 YJInDel-324 5.7 YJInDel-321–YJInDel-324 5.1–5.7 4.28 9.3 0.56 Joiku462
2015S 3.0 YJInDel-324 5.7 YJInDel-324–YJInDel-340 5.7–8.6 5.11 14.5 0.65 Joiku462
qPL11 11 2014P 3.0 YJInDel-436 17.0 YJInDel-435–YJInDel-441 12.0–18.8 3.30 9.6 0.43 Yukihikari
qPL12 12 2014S 3.0 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 4.10 11.1 0.68 Yukihikari
PN qPN7 7 2014P 2.9 YJInDel-304 29.0 YJInDel-301–YJInDel-304 27.0–29.0 3.78 11.2 1.09 Joiku462
qPN9 9 2015P 2.8 YJInDel-356 5.7 YJInDel-356–YJInDel-358 5.7–9.4 3.09 10.2 1.59 Yukihikari
qPN12 12 2014P 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 3.63 10.7 1.15 Joiku462
CL qCL6 6 2014P 3.0 YJInDel-218 9.7 YJInDel-208–YJInDel-218 7.9–9.7 4.71 10.0 2.33 Yukihikari
2015P 3.0 YJInDel-218 9.7 YJInDel-208–YJInDel-218 7.9–9.7 6.67 18.4 3.65 Yukihikari
qCL7 7 2014P 3.0 YJInDel-304 29.0 YJInDel-301–YJInDel-304 27.0–29.0 4.88 10.1 2.48 Yukihikari
qCL8 8 2014P 3.0 YJInDel-321 5.1 YJInDel-320–YJInDel-321 4.1–5.1 3.90 8.2 2.10 Joiku462
2014S 2.9 YJInDel-321 5.1 YJInDel-320–YJInDel-321 4.1–5.1 5.37 14.5 3.45 Joiku462
qCL12 12 2014P 3.0 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 7.30 16.4 2.85 Yukihikari
2014S 2.9 YJInDel-515 25.4 YJInDel-510–YJInDel-515 23.6–25.4 5.08 13.7 3.39 Yukihikari
2015P 3.0 YJInDel-515 25.4 YJInDel-515 25.4 4.14 10.9 2.75 Yukihikari
a  Physical position based on the Nipponbare sequence (RAP-DB build 5.0).

Fig. 2

Chromosomal locations of QTLs for eating quality, grain appearance quality and yield related traits in the RILs derived from the cross between Yukihikari and Joiku462. The chromosome number is shown at the top. Vertical bars denote the linkage maps constructed for the RILs (Kinoshita et al. 2016). Map positions of the QTLs are shown to the right of each chromosome. The length of the vertical bars represents the QTL confidence interval (P < 0.05) and the horizontal bars represent the highest LOD score peak. White and black arrows on the top show that Yukihikari and Joiku462 alleles, respectively, increase the respective traits. Abbreviations: 2014P, 2014 Pippu; 2014S, 2014 Sapporo; 2015P, 2015 Pippu; 2015S, 2015 Sapporo; DTH, days to heading; AAC, apparent amylose content; PC, protein content; BGW, brown grain weight per plant; TBGW, 1000 brown grain weight; BGL, brown grain length; BGWI, brown grain width; BGT, brown grain thickness; GN, grain number per plant; GNP, grain number per panicle; FGN, filled grain number per plant; UFG, unfilled grain ratio; PL, panicle length; PN, panicle number; CL, culm length.

Five QTLs for DTH, qDTH3, qDTH6, qDTH7, qDTH8 and qDTH12, were identified on chromosomes 3, 6, 7, 8 and 12, respectively. Two of these QTLs, qDTH6 and qDTH8, had a major impact on phenotypic variation, with qDTH6 accounting for 33.4%, 21.3%, 36.5% and 43.3% of the total phenotypic variation in 2014P, 2014S, 2015P and 2015S, respectively, and qDTH8 accounting for 12.5%, 25.5%, 15.8% and 9.0%, respectively, of these variations. The QTLs qDTH6 and qDTH8 were associated with extended heading dates of Yukihikari and Joiku462 alleles. Two other QTLs, qDTH3 and qDTH12, showed effects in 2014P and 2015S and were associated with extended heading dates of Joiku462 and Yukihikari alleles. An additional minor QTL, qDTH7, showed effects only in 2014P and was associated with an extended heading date in a Yukihikari allele.

Three QTLs for AAC, qAAC4, qAAC8 and qAAC9 were identified on chromosomes 4, 8 and 9, respectively, and detected across all four environments. The QTL qAAC9 had a major impact on phenotypic variation, accounting for 75.6%, 71.6%, 71.9% and 73.3% of the total phenotypic variations in 2014P, 2014S, 2015P and 2015S, respectively. This QTL was associated with increased AAC in a Yukihikari allele. Two other two QTLs, qAAC4 and qAAC8, had minor effects in the four environments and were associated with increased AAC in a Joiku462 allele. Eight QTLs for PC, qPC1, qPC2, qPC3, qPC6.1, qPC6.2, qPC8, qPC12.1 and qPC12.2, were identified on chromosomes 1, 2, 3, 6 (two QTLs), 8, and 12 (two QTLs), respectively. Two of these QTLs, qPC6.2 and qPC8, had a major impact, with qPC6.2 accounting for 22.3% and 24.4% of the phenotypic variation in 2014P and 2015P, respectively, and qPC8 accounting for 10.5%, 13.3% and 8.5% of the phenotypic variations in 2014P, 2014S and 2015P, respectively. The QTLs qPC6.2 and qPC8 were associated with increased PC in a Joiku462 and a Yukihikari allele, respectively. Six additional QTLs for PC, qPC1, qPC2, qPC3, qPC6.1, qPC12.1 and qPC12.2, were detected in a single environment. The Joiku462 alleles at qPC1 and qPC12.2 increased PC, whereas the Yukihikari allele at qPC2, qPC3 and qPC12.1 increased PC.

Six QTLs for BGW were identified on chromosomes 1, 6 (two QTLs), 8, 9 and 12. The two QTLs on chromosome 6, qBGW6.1 and qBGW6.2, were detected in 2014P and 2015P, respectively, and were associated with increased BGW in a Joiku462 and a Yukihikari allele, respectively. Two additional QTLs, qBGW8 and qBGW12, were detected in 2014S and 2015P, respectively, and were QTLs associated with increased BGW in a Joiku462 and a Yukihikari allele, respectively. The other two QTLs, qBGW1 and qBGW9, were detected in 2014P and 2015P and were associated with increased BGW of Yukihikari alleles.

Seven QTLs for TBGW were identified on chromosomes 1, 4, 8, 9, 11 and 12. The QTL qTBGW1 was detected at both locations in 2015 and explained 6.9% and 9.7% of the total phenotypic variations in 2015P and 2015S, respectively. The QTL qTBGW4 was detected across all four environments and accounted for 13.0%, 15.3%, 13.1% and 15.8% of the total phenotypic variations in 2014P, 2014S, 2015P and 2015S, respectively. The QTL qTBGW8 was detected in 2014P and 2015P, accounting for 9.1% and 5.3%, respectively, of the total phenotypic variation. The QTL qTGBW9 accounted for 10.9%, 8.8% and 7.8%, of total phenotypic variations in 2014P, 2014S and 2015P, respectively, whereas the QTL qTBGW11 accounted for 12.9%, 9.6% and 6.6%, respectively, of the total phenotypic variations in these environments. The two QTLs on chromosome 12, qTBGW12.1 and qTBGW12.2, explained 5.9% and 8.7% of the total phenotypic variations in 2014P and 2015P, respectively. TBGW was increased by the Joiku462 alleles at qTBGW1, qTBGW4, qTBGW8 and qTBGW9, and by the Yukihikari alleles at qTBGW11, qTBGW12.1 and qTBGW12.2.

Two major QTLs for BGL, qBGL4 and qBGL11, were identified on chromosomes 4 and 11, respectively. Both were detected across all four environments, with qBGL4 accounting for 22.3%, 17.9%, 18.3% and 18.8% of the total phenotypic variations in 2014P, 2014S, 2015P and 2015S, respectively, and qBGL11 accounting for 20.4%, 21.0%, 16.3% and 22.9%, respectively, of the phenotypic variations in these environments. BGL was increased by the Joiku462 allele at qBGL4 and by the Yukihikari allele at qBGL11. Four QTLs for BGWI, qBGWI8, qBGWI9, qBGWI12.1 and qBGWI12.2, were identified on chromosomes 8, 9 and 12 (two QTLs), respectively. The QTL qBGWI8 was detected in 2015P, accounting for 7.5% of the total phenotypic variation. In contrast, the QTL qBGWI9 was detected across four environments, accounting for 10.9%, 8.4%, 12.5% and 9.8% of the total phenotypic variations in 2014P, 2014S, 2015P and 2015S, respectively. Of the two QTLs on chromosome 12, one, qBGWI12.1, was detected in 2014P and 2014S and accounted for 10.1% and 11.5%, respectively of the total phenotypic variation in these environments. The second QTL, qBGWI12.2, was detected in 2015P and 2015S and accounted for 9.7% and 11.7%, respectively, of the total phenotypic variation in these environments. BGWI was increased by the Joiku462 alleles at qBGWI8 and qBGWI9 and by the Yukihikari alleles at qBGWI12.1 and qBGWI12.2.

Seven QTLs for BGT were identified, on chromosomes 1, 4 (two QTLs), 7, 8, 9 and 12. The QTL qBGT1 on chromosome 1 was detected across all four environments, accounting for 6.4%, 7.0%, 12.6% and 12.0% of total phenotypic variations in 2014P, 2014S, 2015P and 2015S, respectively. One QTL on chromosome 4, qBGT4.1, was detected in 2014S and 2014S and accounted for 9.7% and 11.0%, respectively, of total phenotypic variations in these environments. The other QTL on chromosome 4, qBGT4.2, was detected in 2015S and 2015P, accounting for 13.1% and 14.6%, respectively, of total phenotypic variations in these environments. Three QTLs, qBGT7, qBGT8 and qBGT9, located on chromosomes 7, 8 and 9, respectively, were detected in 2014P, accounting for 7.9%, 7.9% and 12.7%, respectively, of total phenotypic variations. The QTL qBGT12 on chromosome 12 was detected in 2014P and 2014S, accounting for 7.9% and 8.5%, respectively, of total phenotypic variations in these environments. Increased BGT was associated with the Joiku462 alleles at qBGT1, qBGT4.2, qBGT7, qBGT8 and qBGT9 and with the Yukihikari alleles at qBGT4.1 and qBGT12.

Five QTLs for GN were identified on chromosomes 2, 3, 6, 8 and 9. The QTLs qGN2, qGN3 and qGN8 were detected in 2014S, whereas qGN6 and qGN9 were detected in 2014P. These QTLs accounted for 8.5–12.1% of total phenotypic variation in these environments. Increased GN was associated with the Joiku462 alleles at qGN2, qGN3 and qGN8 and the Yukihikari alleles at qGN6 and qGN9. Five QTLs were also identified for GNP on chromosomes 4, 6, 7, 8 and 12. The QTLs qGNP6 and qGNP12 were detected consistently across the four environments, accounting for 8.5–17.6% of total phenotypic variation in these environments. The QTLs qGNP4 and qGNP7 were detected in 2014P and qGNP8 in 2014S, with these three accounting for 6.9%, 9.7% and 11.9%, respectively, of total phenotypic variations in these environments. Increased GNP was associated with the Yukihikari alleles at qGNP4, qGNP6, qGNP7 and qGNP12 and with the Joiku462 allele at qGNP8.

Six QTLs for FGN were identified on chromosomes 1, 2, 3, 6, 8 and 9. Three QTLs, qFGN1, qFGN6 and qFGN9, were detected in 2014P, and the other three, qFGN2, qFGN3 and qFGN8, in 2014S. Each QTL accounted for 8.9–13.7% of the total phenotypic variations in these environments. Increased FGN was associated with the Joiku462 alleles at qFGN1, qFGN6 and qFGN9 and with the Yukihikari alleles at qFGN2, qFGN3 and qFGN8. A single QTL for UFG, qUFG6, was identified on chromosome 6. This QTL was detected in 2015P and 2015S and accounted for 12.7% and 9.7%, respectively, of the total phenotypic variations in these environments. Increased UFG was associated with the Joiku462 allele at this QTL.

Six QTLs for PL were identified, on chromosomes 4, 6 (two QTLs), 8, 11 and 12. The QTL qPL4 on chromosome 4 was detected in 2015P, accounting for 7.1% of total phenotypic variation. One of the QTLs on chromosome 6, qPL6.1, was detected in 2015P and 2015S, accounting for 16.2% and 17.3%, respectively, of total phenotypic variations. The second QTL, qPL6.2, was detected in 2014P and 2015P, accounting for 13.2% and 8.2%, respectively, of total phenotypic variations. The QTL qPL8 on chromosome 8 was detected in 2014S, 2015P and 2015S, accounting for 15.0%, 9.3% and 14.5%, respectively, of total phenotypic variations. The QTLs qPL11 and qPL12 were detected in 2014P and 2014S, accounting for 9.6% and 11.1%, respectively, of total phenotypic variations. Increased PL was associated with the Joiku462 alleles at qPL4 and qPL8 and with the Yukihikari alleles at qPL6.1, qPL6.2, qPL11 and qPL12.

Three QTLs for PN were identified on chromosomes 7, 9 and 12. The QTLs qPN7 and qPN12 were both detected in 2014P, accounting for 11.2% and 10.7%, respectively, of total phenotypic variations. The third QTL, qPN9, was detected in 2015P and accounted for 10.2% of total phenotypic variation. Increased PN was associated with the Joiku462 alleles at qPN7 and qPN12 and the Yukihikari allele at qPN9. Four QTLs for CL were identified on chromosomes 6, 7, 8 and 12. The QTL qCL6 was detected in 2014P and 2015P, accounting for 10.0% and 18.4%, respectively, of total phenotypic variations, whereas qPL7 was detected in 2014P, accounting for 10.1% of total phenotypic variation. The QTL qCL8 was detected in 2014S and 2014P, accounting for 16.4% and 13.7%, respectively, of total phenotypic variations. An additional QTL, qCL12, was detected in 2014P, 2014S and 2015P, accounting for 16.4%, 13.7% and 10.9%, respectively, of total phenotypic variations. Increased CL was associated with the Joiku462 allele at qCL8 and with the Yukihikari alleles at qCL6, qCL7 and qCL12.

Chromosomal regions associated with multiple QTLs

Fifteen intervals on chromosomes 1, 2, 3, 4, 6, 7, 8, 9, 11 and 12 were found to harbor multiple QTLs affecting the different traits (Table 2, Fig. 2). The distal region of the short arm of chromosome 1 between YJInDel-9 and YJInDel-12 (7.1–8.7 Mb) was observed to have effects on FGN and TBGW. Another region on chromosome 1, between YJInDel-34 and YJInDel-536_2 (26.2–29.4 Mb), controlled BGW and PC. The region on chromosome 2 between YJInDel-65 and YJInDel-67 (9.9–13.5 Mb) was found to affect PC, GN and FGN. On chromosome 3, the region around YJInDel-127 (22.0 Mb) on chromosome 3 was found to control GN and FGN, while the region around YJInDel-129 (24.5 Mb) was found to affect DTH and PC. The regions near YJInDel-161 (28.0 Mb) and YJInDel-162 (29.2 Mb) on chromosome 4 had effects on TBGW, BGL and BGT. Eleven QTLs on chromosome 6 were observed to cluster in two regions. The region between YJInDel-207 and YJInDel-218 (5.2–9.7 Mb) included nine QTLs, qDTH6, qPC6.2, qBGW6.2, qGN6, qGNP6, qFGN6, qUFG6, qPL6 and qCL6. An additional two QTLs, qPC6.1 and qBGW6.1, were clustered near YJInDel-197 (0.8 Mb) at the distal region of the short arm of chromosome 6. The distal region on the long arm of chromosome 7 between YJInDel-301 and YJInDel-304 (27.0–29.0Mb) was associated with DTH, GNP, PN and CL. Eleven QTLs on chromosome 8 were observed to cluster in two regions, with the region between YJInDel-306 and YJInDel-324 (3.0–5.4 Mb) including qDTH8, qAAC8, qPC8, qBGT8, qGN8, qGNP8, qPL8 and qCL8; and the region between YJInDel-340 and YJIInDel-341 (8.6–17.0 Mb) including qBGW8 and qTBGW8. Eight QTLs on chromosome 9, qAAC9, qBGW9, qTBGW9, qBGWI9, qBGT9, qGN9, qFGN9 and qPN9, were found to cluster in the distal region of the short arm between YJInDel-351 and YJInDel-358 (3.7–9.4 Mb). The region of chromosome 11 between YJInDel-435 and YJInDel-442 (12.0–19.8 Mb) was associated with TBGW and BGL. Twelve QTLs on chromosome 12 were found to cluster in two flanking regions on the long arm. The region between YJInDel-502 and YJInDel-504 (19.3–21.6 Mb) harbored three QTLs, qTBGW12.1, qBGWI12.1 and qBGT12.1, and the region near YJInDel-515 (25.4 Mb) harbored nine QTLs, qDTH12, qPC12.2, qBGW12, qTBGW12.2, qBGWI12.2, qGNP12, qPL12, qPN12 and qCL12.

Discussion

Genetic improvements in eating quality

In the present study, we identified a total of 72 QTLs associated with eating quality, grain appearance and yield related traits on 10 rice chromosomes. Based on these findings, along with those of our previous study on the glossiness area (GLA) and glossiness strength (GLS) of cooked rice and the whiteness of polished rice (WPR) (Kinoshita et al. 2016), we summarized the QTLs on rice chromosomes (Fig. 3). Yukihikari was found to have two minor QTLs for reduced AAC, qAAC4 and qAAC8, whereas Joiku462 had a major QTL for qAAC9. This qAAC9 allele from Joiku462 had been detected at a similar chromosomal region (Shinada et al. 2015). Using marker-assisted selection (MAS), qAC9.3 (Ando et al. 2010) was introduced into Joiku462 from Hokkai PL9 (Shinada et al. 2015), suggesting that qAAC9 should be identical to qAC9.3. The present study showed that a high proportion of total phenotypic variation (>70%) could be explained by qAAC9 across the four environmental conditions. The QTL qAAC8 was found to be located at the same interval as qDTH8 or an adjacent interval, with the early heading allele at qDTH8 from Yukihikari combined with the reduced AAC allele at qAAC8, also from Yukihikari. The QTL cluster for AAC and DTH was previously reported located in a similar region on chromosome 8 (Kwon et al. 2011, Yamamoto et al. 1998, Wan et al. 2004). The Yukihiari allele at qAAC4 has a small impact on reduced AAC without modifying DTH. To test whether qAAC4, when combined with qAAC9, which had no effect on DTH over several years at different locations, would be useful in a breeding program for the fine-tuning of AAC without modifying DTH, we are developing near isogenic lines (NILs) for each combination at the two QTLs.

Fig. 3

Distribution of QTLs for eating quality, grain appearance quality and yield related traits on rice chromosomes. The chromosome number is shown at the top. The black boxes indicate the positions of the QTL confidence intervals (P < 0.05) by physical distance. Up and down arrows indicate that traits are enhanced and reduced, respectively, by Joiku462 alleles. The font size of the QTL designation indicates QTL stability, with the largest font indicating that the QTL was detected in all four environments, the larger font indicating that the QTL was detected in three environments, the smaller font indicating that the QTL was detected in two environments and the smallest font indicating that the QTL was detected in one environment. QTLs for the appearance of cooked rice and polished rice had been identified by Kinoshita et al. (2016) and are represented by asterisks. Abbreviations: GLA, glossiness area of cooked rice; GLS, glossiness strength of cooked rice; WPR, whiteness of polished rice.

Wx/GBSSI gene expression was lower in Joiku462 than in Yukihikari, despite both harboring the Wxb allele (Takano et al. 2014a). In addition, the amount of Wx/GBSSI protein was reduced by qAC9.3 (Ando et al. 2010), suggesting that qAC9.3/qAAC9 may reduce Wx expression at both the transcriptional and post-transcriptional levels. The NILs described above may also be useful in studying the ability of qAC9.3/qAAC9 and qAAC4 to regulate AAC. Thus, this study found that Yukihikari had accumulated two minor QTLs, qAAC8 and qAAC4, derived from old Hokkaido landraces, whereas Joiku462 had gained a major QTL, qAC9.3/qAAC9, while eliminating qAAC4 and qAAC8 from Yukihikari.

Joiku462 has been shown to reduce PC relative to Yukihikari (Shinada et al. 2015). The present study revealed a complex genetic system controlling PC in Joiku462 and Yukihikari. Joiku462 was found to have the allele for reduced PC at five QTLs, qPC2, qPC3, qPC6.1, qPC8 and qPC12.2, whereas Yukihikari had the allele for reduced PC at three QTLs, qPC1, qPC6.2 and qPC12.1. Joiku462 has been reported to have a QTL at chromosomal position similar to that of qPC3 (Shinada et al. 2015). The Joiku462 allele in this region, however, increased PC in the Joiku462/Joukei06214 double haploid population (Shinada et al. 2015). In addition, qPC2 in this study was detected at a different position on chromosome 2 than the previously described PC QTL (Shinada et al. 2015). Taken together, these findings indicate that the QTLs qPC2 and qPC3 identified in the current study are distinct from those of the previous study. Moreover, seven of the eight QTLs for PC were detected in QTL clusters. Two major QTLs for PC, qPC6.2 and qPC8, are combined with two QTLs for DTH, qDTH6 and qDTH8, whereas two minor QTLs for PC, qPC3 and qPC12.2, are combined with qDTH3 and qDTH12. At all four QTL clusters, late heading alleles combined with those for reduced PC. In addition, qPC6.2, qPC8 and qPC12.2 were located within the same intervals or those flanking multiple QTLs for BGW, GN, FGN, GNP, PL and CL on their respective chromosomes. At these three QTL clusters, alleles for reduced PC were linked to alleles for increased BGW, GN, FGN, GNP, PL and/or CL. In addition, qPC8 was found to be located within the same or a flanking interval of qAAC8, with the allele for reduced PC linked with the allele for increased AAC. Two PC QTLs, qPC1 and qPC6.1, were located within the same intervals as each of two BGW QTLs, qBGW1 and qBGW6.1, respectively. In both QTL clusters, the allele for high yield was linked with the allele for reduced PC. The QTL qPC2 was located in the same interval as QTLs for GN and FGN, qGN2 and qFGN2, respectively. Within this QTL cluster, the Joiku462 alleles for increased GN and FGN were combined with those for reduced PC. These findings of QTL clusters for PC and yield-related traits were consistent with negative correlations between phenotypic variables. Protein accumulation in rice grains is associated with nitrogen uptake and nitrogen flow dynamics within the plant. The QTLs for PC identified in this study can be classified into four groups: (1) qPC1, qPC2 and qPC6.1, which are associated with the secondary effects of large biomasses (sinks and/or sources); (2) qPC3, which is associated with the secondary effects of temperature during the filling period through the modification of DTH; (3) qPC6.2, qPC8 and qPC12.2, which are associated with the combined secondary effects of large biomasses and temperature; and (4) qPC12.1, the QTL for PC itself. Further study is required to determine the precise positions of QTLs for PC and multiple traits in QTL clusters, as well as to assess whether pleiotropic effects are due to a single or closely linked QTLs. Nevertheless, qPC12.1 was shown to contribute to reduced PC, independent of other traits in Joiku462. To our knowledge, there is no other QTL/gene for PC at chromosomal position similar to that of qPC12.1 (Q-TARO database, Yonemaru et al. 2010). qPC12.1 should therefore be evaluated during multiple years and locations using precise genetic stocks such as NILs. Further improvements in PC of Joiku462 may require the introduction of a novel major gene/QTL for PC.

Genetic improvements in grain appearance quality

Compared with Yukihikari, Joiku462 has shown improved grain appearance, involving BGL, BGWI and BGT. Especially, Joiku462 yielded stable thick brown rice grains of thickness >2.0 mm across all four environments, whereas Yukihikari yielded grains <2.0 mm at both locations in 2014. The present study identified 13 QTLs for grain shape, including eight for BGT, four for BGWI and two for BGL. Yukihikari had five QTLs, qBGT4.1, qBGT12, qBGWI12.1, qBGWI12.2 and qBGL11, for increased BGWI or BGT. In contrast, Joiku462 had eight QTLs, qBGT1, qBGT4.2, qBGT7, qBGT8, qBGT9, qBGWI8, qBGWI9 and qBGL4, for increased BGL, BGWI or BGT. Six QTLs, qBGL4 and qBGT4.2 on chromosome 4, qBGWI9 and qBGT9 on chromosome 9, and qBGWI12.2 and qBGT12 on chromosome 12, were found to be located within the same or adjacent intervals. Joiku462 had two QTL clusters, qBGT4.2qBGL4 and qBGT9qBGWI9, for increased BGL, BGWI or BGT, while Yukihikari had a QTL cluster, qBGT12qBGWI12.2, for increased BGWI or BGT, suggesting that each of these combinations of traits was related, findings consistent with positive correlations between phenotypic variables. Five QTLs/QTL clusters, qBGL11 and qBGWI12.2 from Yukihikari and qBGT1, qBGWI9 and qBGT4.2qBGL4 cluster from Joiku462, were detected repeatedly in three or more environments, whereas the remaining QTLs were detected in fewer than three environments. No QTL for BGL and/or BGWI was located in the qBGT1 region, suggesting that qBGT1 has a genetic mechanism different from other QTLs for BGT.

Taken together, these findings showed that improvements of BGT in Joiku462 were due to the introgression in Yukihikari of a series of QTLs, two stable QTLs, qBGT1 and qBGT4.2, and three unstable QTLs, qBGT7, qBGT8 and qBGT9. At present, we are developing NILs for each QTL and planning to test each under various environmental conditions. The molecular basis of BGT control will also be clarified using NILs for each BGT QTL.

Potential to increase grain yield

In the present study, BGW of the parental cultivars did not differ significantly under four environmental conditions. In the RIL population, however, transgressive segregants with extremely high and low yields were observed in each environment, suggesting that genetic variations underlying BGW were segregated in the RIL population. The RILs showed stable positive correlations between BGW and other traits, including DTH, GNP, FGN, PL and CL, across the four environmental conditions. These findings suggested that increasing GNP or FGN with relatively longer PL, combined with relatively longer CL and later DTH, contributed to increased BGW. Significant positive correlations between BGW and BGWI were observed in three of the four environments, suggesting that wide grains contribute to high BGW. In addition, significant positive correlations between BGW and TBGW were observed in Pippu in both years, suggesting that the contribution of TBGW to increased BGW is specific to Pippu.

The present study identified 42 QTLs for yield related traits, including six for BGW, seven for TBGW, five for GN, five for GNP, six for FGN, six for PL, three for PN, and four for CL. All six QTLs for BGW were located in QTL clusters for other trait(s) and could be classified into three groups based on combined trait(s). Two QTLs, qBGW1 and qBGW6.1, were located in the same or adjacent interval as qPC1 and qPC6.1, respectively. Within both clusters, QTLs for increased BGW were combined with QTLs for reduced PC from each parent, qPC1 from Yukihikari and qPC6.1 from Joiku462. Three other QTLs for BGW, qBGW6.2, qBGW8 and qBGW12, were located in the same or adjacent intervals as QTLs qDTH6, qDTH8 and qDTH12, respectively, for DTH. At all three QTL clusters, increased BGW was combined with late heading. In addition, the qBGW6.2qDTH6 cluster included qGN6, qGNP6, qCL6 and qPL6.1, with the allele for high yield combined with the alleles for increases in each other trait. Similarly, the qBGW8qDTH8 cluster included qGN8, qCL8 and qPL8, with the allele for high yield again combined with the alleles for increases in each other trait. Furthermore, the qBGW12qDTH12 cluster included qTBGW12, qBGWI12.2, qGNP12, qPN12, qPL12 and qCL12, with alleles for high yield again combined with alleles for increases in each other trait. Taken together, these findings indicate that large biomass and increased GN/GNP contributed to increased yields through the qBGW6.2qDTH6 and qBGW8qDTH8 clusters; and that increased BGWI leading to increased TBGW contributed to higher yield through the qBGW12qDTH12 cluster. It should be noted that later heading alleles at all three QTL clusters were combined with low PC alleles. Finally, qBGW9 was located in a complex QTL cluster, containing not only qTBGW9, qBGWI9, qBGT9, qFGN9 and qPN9, but also qAAC9 and GLS9 (Kinoshita et al. 2016). In addition to low AAC and high GLS, the Joiku462 allele was associated with increases in BGW, TBGW, BGWI and BGT, despite reductions in GN, FGN and PN. AAC correlated negatively with the glossiness of cooked rice (Juliano et al. 1965, Takeuchi et al. 2007, Tanaka et al. 2006). Furthermore, QTLs for glossiness were mapped to approximately the same regions on chromosomes 2, 3, and 6 as QTLs for amylose content (Takeuchi et al. 2007, Tanaka et al. 2006). In addition, the results of the present study support those of our previous study on GLS (Kinoshita et al. 2016), which found that reduced AAC contributed to the increased glossiness of cooked rice. In contrast, to our knowledge, increased grain size has not been previously reported to reduce amylose content. Therefore, this complex QTL cluster is a key determinant not only of eating quality but of controlling the balance between grain number and grain size for the improvements of yield. Fine mapping of each QTL using an advanced backcross population may clarify whether closely linked genes or the pleiotropic effect of a single locus contributed to these QTL clusters. In addition, further studies are required to determine the usefulness of the two clusters, the qBGW1qPC1 cluster from Yukihikari and the qBGW6.1qPC6.1 cluster from Joiku462, in future rice breeding programs.

Acknowledgments

The authors thank Drs H. Miura, K. Onishi, M. Mori and J. Kasuga for their valuable comments throughout this study. The authors also thank H. Nagasawa, H. Yamamoto, M. Sato, and part-time research assistants at the Kamikawa Agricultural Experiment Station for their assistance in field experiments. Paddy field experiments were performed at the Kamikawa Agricultural Experiment Station, Agricultural Research Department, Hokkaido Research Organization and the Field Science Center for Northern Biosphere, Hokkaido University. This study was supported by the Tojuro Iijima Foundation for Food Science and Technology (TS and KK).

Literature Cited
 
© 2017 by JAPANESE SOCIETY OF BREEDING
feedback
Top