Distribution of different Vrn-B1 alleles in hexaploid spring wheat germplasm

In wheat, the transition from the vegetative to reproductive stage is primarily controlled by the series of vernalisation (Vrn-1) genes located on the homoeologous group 5 chromosomes. Up to 2009, only two alleles at the Vrn-B1 locus were known: one dominant, spring, allele (now designated Vrn-B1a) and the other recessive, winter, (vrn-B1) allele. Recently, two additional dominant alleles, Vrn-B1b and Vrn-B1c, were described. In this study, we screened a range of hexaploid spring wheat germplasms for the presence of different Vrn-B1 alleles using new diagnostic molecular markers. Our results show that the Vrn-B1a allele was the most prevalent, being present in 55.3 % of the 2,495 accessions examined, followed by the recessive vrn-B1 allele, which occurred in 31.5 % of the accessions. The novel alleles Vrn-B1b and Vrn-B1c were found in 5.3 and 7.9 % of all accessions, respectively.     

Relative performance of Canadian spring wheat cultivars under organic and conventional field conditions

Producing higher yields under organic conditions is generally hampered by weeds and lesser nutrient supply. In wheat certain adaptive traits like early season vigour, taller plants, and shorter life cycle have been reported to help plants compete with weeds and produce satisfactory yields. In this experiment we tested the hypothesis ‘that early flowering and maturity conferred by insensitive vernalization alleles Vrn-A1a and/or Vrn-B1 has a yield advantage under organic conditions’ in Canadian spring wheat germplasm. We genotyped 32 cultivars for their vernalization gene composition (Vrn-A1a, Vrn-B1 and Vrn-D1) and studied these cultivars in organic and conventional management systems. We found 88 % of the cultivars possessed vernalization (Vrn) insensitive allele Vrn-A1a either alone or in combination with Vrn-B1. There were no differential affects between the cultivars having insensitive Vrn allele at either single locus (Vrn-A1a) or two (Vrn-A1a, Vrn-B1) under organic and conventional field conditions; except for days to maturity, where cultivars having only Vrn-A1a allele matured earlier. This earlier maturity did not translate to any yield advantage under organic field conditions. Overall, the cultivars grown under organic conditions were earlier flowering, lower yielding with lower test weight compared to the conventional management system. Significant cultivar × environment interactions were found for grain yield, grain protein content and grain fill rate. For grain protein content, cross-over interactions of the cultivars between the management systems were observed. Three cultivars (Marquis, Unity and Minnedosa) exhibited minimal comparative loss in grain yield and grain protein content under organic field conditions, and hence could potentially serve as parents for organic wheat breeding programs.     

Analysis of contributors to grain yield in wheat at the individual quantitative trait locus level

In wheat, strong genetic correlations have been found between grain yield (GY) and tiller number per plant (TN), fertile spikelet number per spike (FSN), kernel number per spike (KN) and thousand‐kernel weight (TKW). To investigate their genetic relationships at the individual quantitative trait locus (QTL) level, we performed both normal and multivariate conditional QTL analysis based on two recombinant inbred lines (RILs) populations. A total of 79 and 48 normal QTLs were identified in the International Triticeae Mapping Initiative (ITMI)/SHW‐L1 × Chuanmai 32 (SC) populations, respectively, as well as 55 and 35 conditional QTLs. Thirty‐two QTL clusters in the ITMI population and 18 QTL clusters in the SC population explained 0.9%–46.2% of phenotypic variance for two to eight traits. A comparison between the normal and conditional QTL mapping analyses indicated that FSN made the smallest contribution to GY among the four GY components that were considered at the QTL level. The effects of TN, KN and TKW on GY were stronger at the QTL level.     

Temperature as a component of the expression of developmental responses in wheat

Summary Studies were made of days to ear emergence under the constant temperatures of 9, 14, 19 and 25°C and 16 h photoperiod in three sets of wheat lines each possessing genotypes differing for developmental responses.Days to ear emergence in three near-isogenic lines of the wheat cultivar Triple Dirk, which differed for vernalization response, increased as the strength of the response increased. At the four temperatures Triple Dirk D (Vrn 1 vrn 2) was not significantly different from normal Triple Dirk (Vrn 1 Vrn 2) but Triple Dirk B (vrn 1 Vrn 2) was significantly (P=0.01) later than normal Triple Dirk at each temperature. This indicates that the vrn 1 allele confers stronger vernalization response than vrn 2 over the range of temperatures (9–24°C). However, Triple Dirk C (vrn 1 vrn 2) failed to head after 120 days at each temperature indicating strong interaction between vrn 1 and vrn 2 with each other (and possibly the Triple Dirk back-ground) to give a much stronger vernalization response than predictions from additivity of their individual effects.The second set comprised the four Chinese Spring/Thatcher chromosome substitution lines CS/T 3B, 6B, 7B and 5D, plus Chinese Spring and Thatcher, and were grown in the unvernalized condition. CS/T 5D was similar in days to ear emergence as Chinese Spring at all four temperatures but the other three lines were earlier to ear emergence, particularly as the temperature increased. Days to ear emergence was fastest at 14°C in all lines, except CS/T 3B, in which it decreased progressively from 9 to 24°C.The third set of Chinese Spring and Thatcher and the homoeologous group 2 chromosomes of Thatcher substituted in Chinese Spring, the group which is considered to be involved in the control of photoperiod sensitivity. The three substitution lines responded differently to temperature compared with Chinese Spring and with each other, with chromosome 2D being the least, and chromosome 2B the most, responsive to temperature.     

QTL mapping of yield-related traits in the wheat germplasm 3228

The new wheat germplasm 3228, a putative derivative of tetraploid Agropyron cristatum Z559 and the common wheat Fukuhokomugi, has superior features in yield-related traits, particularly in spike morphological traits, such as large spike and superior grain number. To identify favorable alleles of these traits in 3228, 237 F_2:3 families were developed from the cross 3228/Jing 4839. A genetic map was constructed using 179 polymorphic SSR and EST-SSR markers. A total of 76 QTL controlling spike number per plant (SNP), spike length (SL), spikelet number per spike (SNS), floret number per spikelet (FNS), grain number per spike (GNS) and thousand-grain weight (TGW) were detected on 16 chromosomes. Each QTL explained 1.24–27.01% of the phenotypic variation, and 9 QTL (28.95%) were detected in two or all environments. Additive effects of 45 QTL were positive with 3228 alleles increasing the QTL effects, 31 QTL had negative effects indicating positive contributions from Jing 4839. Three important clusters involving all traits were located on chromosomes 5A, 6A and 4B, and several co-located QTL were also found. Most of the QTL detected on the three chromosome regions could contribute to the use of 3228 in breeding for grain yield improvement.     

Upland rice breeding in Brazil: a simultaneous genotypic evaluation of stability, adaptability and grain yield

A durum wheat recombinant inbred line population developed from PDW 233 × Bhalegaon 4 cross was analyzed in five environments to understand the genetic network responsible for test weight (TW), thousand kernel weight (TKW), grain yield (YLD), spike length (SL), spikelets per spike (SPS), kernels per spike (KER) and kernel weight per spike (KWS). Genotype, environment and their interactions were main sources of variance for all the traits. TW and TKW were influenced by 11 main effect QTL and 6 digenic epistatic interactions detected on chromosomes 2A, 2B, 4B and 7A. Grain yield was influenced by three epistatic interactions and five main effect QTL, of which two on chromosome 2A were most consistent. A major QTL for spike length was observed on chromosome 3B. QTL for spike characters were distributed over 9 chromosomes. All the traits showed significant influence of digenic epistasis (QQ) and, to a certain extent, QTL × environment interactions (QQE). Therefore, while breeding for complex traits like kernel characters and grain yield components, these interactions should also be considered important. The consistent QTL on chromosome 2A between the marker interval Xgwm71.2–Xubc835.4 with pleiotropic effect on TW and TKW, may be utilized in early generation selection to improve TW and TKW and thereby the milling potential of the durum wheat.     

QTL mapping and analysis of epistatic interactions for grain yield and yield-related traits in <Emphasis Type="Italic">Triticum turgidum</Emphasis> L. var. <Emphasis Type="Italic">durum</Emphasis>

A quantitative trait loci (QTL) analysis of grain yield and yield-related traits was performed on 93 durum wheat recombinant inbred lines derived from the cross UC1113 × Kofa. The mapping population and parental lines were analyzed considering 19 traits assessed in different Argentine environments, namely grain yield, heading date, flowering time, plant height, biomass per plant, and spikelet number per ear, among others. A total of 224 QTL with logarithm of odds ratio (LOD) ≥ 3 and 47 additional QTL with LOD > 2.0 were detected. These QTL were clustered in 35 regions with overlapping QTL, and 12 genomic regions were associated with only one phenotypic trait. The regions with the highest number of multi-trait and stable QTL were 3BS.1, 3BS.2, 2BS.1, 1BL.1, 3AL.1, 1AS, and 4AL.3. The effects of epistatic QTL and QTL × environment interactions were also analyzed. QTL putatively located at major gene loci (Rht, Vrn, Eps, and Ppd) as well as additional major/minor QTL involved in the complex genetic basis of yield-related traits expressed in Argentine environments were identified. Interestingly, the 3AL.1 region was found to increase yield without altering grain quality or crop phenology.     

Pleiotropic effects of the elongated glume gene P1 on grain and spikelet shape-related traits in tetraploid wheat

Wheat grain size and shape are associated not only with yield but also with product and milling quality. A subspecies of cultivated tetraploid wheat, Triticum turgidum ssp. polonicum, is characterized by elongated glumes. To elucidate morphological effects of the subspecies differentiation-related gene, we conducted QTL analysis for grain and spikelet shape using a mapping population between two tetraploid wheat subspecies, polonicum and durum. P1, the gene controlling the elongated glumes, was located on chromosome 7A, and the polonicum-type allele acted in an incomplete dominance manner to express the elongated glume phenotype. The polonicum allele of the P1 locus significantly affected not only glume length but also grain shape, spike shape, awn length and seed fertility in tetraploid wheat. The elongated glume phenotype was correlated with an increase in spike length, grain length and grain weight, and with a decrease in fertility, grain number and awn length. Thus, the subspecies differentiation-related gene in subspecies polonicum dramatically affects grain shape accompanied by alteration of spikelet shape in tetraploid wheat.     

QTL for flag leaf size and their influence on yield-related traits in wheat

Flag leaf-related traits (FLRTs) are determinant traits affecting plant architecture and yield potential in wheat (Triticum aestivum L.). In this study, three related recombinant inbred line (RIL) populations with a common female parent were developed to identify quantitative trait loci (QTL) for flag leaf width (FLW), length (FLL), and area (FLA) in four environments. A total of 31 QTL were detected in four environments. Two QTL for FLL on chromosomes 3B and 4A (QFll-3B and QFll-4A) and one for FLW on chromosome 2A (QFlw-2A) were major stable QTL. Ten QTL clusters (C1–C10) simultaneously controlling FLRTs and yield-related traits (YRTs) were identified. To investigate the genetic relationship between FLRTs and YRTs, correlation analysis was conducted. FLRTs were found to be positively correlated with YRTs especially with kernel weight per spike and kernel number per spike in all the three RIL populations and negatively correlated with spike number per plant. Appropriate flag leaf size could benefit the formation of high yield potential. This study laid a genetic foundation for improving yield potential in wheat molecular breeding programs.     

QTL detection of seven spike-related traits and their genetic correlations in wheat using two related RIL populations

Spike-related traits contribute greatly to grain yield in wheat. To localize wheat chromosomes for factors affecting the seven spike-related traits??i.e., the spike length (SL), the basal sterile spikelet number (BSSN), the top sterile spikelet number (TSSN), the sterile spikelet number in total (SSN), the spikelet number per spike (SPN), the fertile spikelet number (FSN) and the spike density (SD)??two F_8:9 recombinant inbred line (RIL) populations were generated. They were derived from crosses between Weimai 8 and Jimai 20 (WJ) and between Weimai 8 and Yannong 19 (WY), comprising 485 and 229 lines, respectively. Combining the two new linkage maps and the phenotypic data collected from the four environments, we conducted quantitative trait locus (QTL) detection for the seven spike-related traits and evaluated their genetic correlations. Up to 190 putative additive QTL for the seven spike-related traits were detected in WJ and WY, distributing across all the 21 wheat chromosomes. Of these, at least nine pairwise QTL were common to the two populations. In addition, 38 QTL showed significance in at least two of the four different environments, and 18 of these were major stable QTL. Thus, they will be of great value for marker assisted selection (MAS) in breeding programs. Though co-located QTL were universal, every trait owned its unique QTL and even two closely related traits were not excluded. The two related populations with a large/moderate population size made the results authentic and accurate. This study will enhance the understanding of the genetic basis of spike-related traits.     

QTL-based analysis of heterosis for number of grains per spike in wheat using DH and immortalized F 2 populations

To map quantitative trait loci (QTL) and heterotic loci (HL) related to grain number per spike (GNS), 168 double haploid (DH) populations derived from Huapei?3?×?Yumai?57 and an immortalized F ₂ population (IF ₂) generated by randomly permutated intermating of these DH populations were investigated. Using inclusive composite interval mapping (ICIM), a total of nine and eight significant QTLs for GNS were detected in three different environments in DH and IF ₂ populations, respectively. QTLs on chromosomes?1A, 2B, 3B, and 6A were observed between two populations. Five QTLs were detected on chromosome?1A. Of these QTLs, QGns1A-1 was a major QTL explaining 31.25?% of phenotypic variation. QGns2B-2 detected on chromosome?2B had the most significant additive effects, explaining 46.75?% of phenotypic variation with the favorable allele contributed by Yumai?57 corresponding to an increase of 5.69?kernels. Mid-parent heterosis of each cross in the IF ₂ population was used to map heterotic quantitative trait loci. A total of 17 HLs were detected. QTLs and HLs on chromosomes?2B and 6A were observed in the IF ₂ population. Three HLs, QHgns1B-2, QHgns2B, and QHgns6A-1, were detected in two environments and expressed stably. These results showed that some intervals on chromosomes?1B, 2B, and 6A play an important role in GNS heterosis in wheat, improving understanding of this phenomenon.     

Identifying AFLP and microsatellite markers for vernalization response gene Vrn-B1 in hexaploid wheat using reciprocal mapping populations

Marker‐assisted selection may be useful for combining specific vernalization response (Vrn) alleles into a single wheat genotype for yield enhancement; however, DNA markers are only available for two of the three genes identified to date. The objectives of this study were to investigate reciprocal effects on days to heading using F₂ populations generated by cross‐hybridizing near‐isogenic lines (NILs) carrying spring (Vrn‐B1; TDB) and winter (vrn‐B1; TDC) alleles, and to identify markers linked to Vrn‐B1 through genetic linkage analysis. Heading data were recorded for 91 and 89 progeny from reciprocal mapping populations TDB/TDC and TDC/TDB, respectively, and significant (P < 0.0001) reciprocal and dominance effects were detected. Among 207 amplified fragment length polymorphisms primer pairs and seven wheat microsatellite markers screened, two and one, respectively, were linked distally to Vrn‐B1 on wheat chromosome 5BL. Microsatellite Xgwm408 was most closely linked to Vrn‐B1 at 3.9 and 1.1 cM in the TDB/TDC and TDC/TDB map, respectively. Reciprocal differences in recombination distances emphasize the importance of female parent choice when generating mapping populations. Molecular markers are now available for three Vrn loci in wheat.     

Identification and validation of a yield-enhancing QTL cluster in rice (Oryza sativa L.)

Improvement of rice grain yield (YD) is an important goal in rice breeding. YD is determined by its related traits such as spikelet fertility (SF), 1,000-grain weight (TGW), and the number of spikelets per panicle (SPP). We previously mapped quantitative trait loci (QTLs) for SPP and TGW using the recombinant inbred lines (RILs) derived from the crosses between Minghui 63 and Teqing. In this study, four QTLs for SF and four QTLs for YD were detected in the RILs. Comparison of the locations of QTLs for these three yield-related traits identified one QTL cluster in the interval between RM3400 and RM3646 on chromosome 3. The QTL cluster contained three QTLs, SPP3a, SF3 and TGW3a, but no YD QTL was located there. To validate the QTL cluster, a BC₄F₂ population was obtained, in which SPP3a, SF3 and TGW3a were simultaneously mapped to the same region. SPP3a, SF3 and TGW3a explained 36.3, 29.5 and 59.0 % of phenotype variance with additive effect of 16.4 spikelets, 6 % SF and 1.8 g grain weight, respectively. In the BC₄F₂ population, though the region has opposite effects on TGW and SPP/SF, a YD QTL YD3 identified in this cluster region can increase 4.6 g grains per plant, which suggests this QTL cluster is a yield-enhancing QTL cluster and can be targeted to improve rice yield by marker aided selection.     

云南旱地小麦不同穗型品种减源缩库与穗部性状的关系

通过剪叶、剪穗处理探讨云南三种穗型的旱地小麦品种的源库关系，结果认为：平地小麦在齐穗后不同剪叶、剪穗处理，其穗部性状变化明显。大穗型品种云麦39不同的剪叶处理其每穗重量和单粒重下降幅度较大，而小穗型品种云麦42和中穗型品种97591与云麦39相比每穗重量和单粒重下降幅度较小；不同剪叶处理的每穗饱粒重也下降，下降的程度依次是云麦39＞97591＞云麦42。大穗型品种的每穗饱粒数下降明显，而小穗型和中穗型品种的下降较少。对于不同的剪穗处理，大穗型品种的单粒重增加幅度较大，而小穗型品种增加幅度较小，中穗型品种的甚至减少。对于不同穗型品种的穗部性状源库结构调整的反应不同表明：其灌浆过程或光合器官间的相互补偿能力不同，这种不同使不同穗型品种对各种环境的适应性有差异，栽培管理应该有所区别。     

The effect of VRN1 genes on important agronomic traits in high‐yielding Canadian soft white spring wheat

For reproductive success, flowering time must synchronize with favourable environmental conditions. Vernalization genes play a major role in accelerating or delaying the time to flowering. We studied how different vernalization (VRN1) gene combinations alter days to flowering and maturity and consequently the effect on grain yield and other agronomic traits. The study focussed on the effect of the VRN1 gene series (Vrn‐A1, Vrn‐B1 and Vrn‐D1) and their combinations. The Vrn gene group Vrn‐A1a, Vrn‐B1, vrn‐D1 was the earliest to flower and mature, while Vrn‐A1b, Vrn‐B1, vrn‐D1 was the latest to flower. Spring wheat lines with vrn‐A1, Vrn‐B1, Vrn‐D1 were the highest yielding and matured at a similar time as those having vernalization genes Vrn‐A1a, Vrn‐B1 and Vrn‐D1. The findings of this study suggest that the presence of Vrn‐D1 has a direct or indirect role in producing higher grain yield. We therefore suggest the introduction of Vrn‐D1 allele into higher‐yielding classes within Canadian spring wheat germplasm.     

Wheat Development as Affected by Radiation at Two Temperatures

A wheat cultivar (Condor) was grown in two experiments (thermal regimes 18/13 and 21/16°C) under low (298 μE m^-2: s^-1) radiation regimes during either an early phase from seedling emergence to terminal spikelet initiation (S₂), a late phase from terminal spikelet initiation to anthesis (S₂), or for the full period from seedling emergence to anthesis (S₁₂), or high (560 μE m^-2s^-1) radiation throughout the growing period (S₁₂) to determine whether developmental events are affected by radiation. The main developmental events considered in this study were the timing of terminal spikelet initiation and anthesis, the final number of leaf and spikelet primordia initiated in the apex and the rate of leaf appearance. Number of grains per spike and culm height were also measured. The duration of each phenophase was not affected by radiation intensity. Temperature affected the rate of wheat development, but the acceleration of development due to temperature during the seedling emergence - terminal spikelet initiation phase only slightly reduced (from 24.8 to 23.2 days). Differences in time from terminal spikelet initiation to anthesis were greater than in the earlier phases, having been the duration reduced from 24.6 to 20.0 days due to high temperature. Associated with the lack of effect of radiation on phasic development and the negligible effect of temperature on the duration of the early phases of development, final Leaf number was practically unchanged in this study by either the radiation level or the growing temperature. Thus, radiation did not affect the rate of leaf initiation. The number of spikelets was affected by neither the treatments nor the thermal environment. The rates of leaf appearance were accelerated by temperature. Radiation, on the other hand, did not significantly alter the rates of leaf appearance in any of the treatments. As expected from many sources in the literature, the number of grains per spike was significantly affected by radiation during the phase from terminal spikelet initiation to anthesis. Due to the lack of significant effects of radiation on the developmental patterns of wheat, the changes in number of grains per spike were due to changes in the number of grains born in each spikelet. The results of the present study were compared with others available in the literature on the effects (or lack of them) of radiation and CO₂ concentration on phasic development, plastochron and phyllochron in wheat to reach the general conclusion that the rate of developmental events in wheat, in contrast to other plants, is almost completely independent of the availability of assimilates, with a possible exception for the Equatorial latitudes.     

Genetic effects of Vrn genes on heading date and agronomic traits in bread wheat

Summary The Vrn1, Vrn2 and Vrn3 genes have different values of effects on heading date and related yield components. The genetic background and environment do not affect the ranking of Vrn genotypes according to earliness within near-isogenic line sets; however, they do influence the level of differences between heading dates of particular genotypes and between effect values, respectively. The frequencies of defined Vrn genotypes in the global set of spring bread wheat cultivars are associated with grain weight per plant predicted on the basis of Vrn gene effects averaged over backgrounds and over environments. Peculiarities of backgrounds and environments alter the grain yield ranges of Vrn genotypes. For early photoperiod-insensitive wheats, planted in stress conditions at grain filling, the highest yield was predicted for double dominant Vrn genotypes with Vrn3. This gene is rarely used by the breeders in middle latitudes and its wider adoption is encouraged.     

QTL analysis for grain weight in common wheat

Quantitative trait loci (QTL) analysis for grain weight (GW = 1000 grain weight) in common wheat was conducted using a set of 100 recombinant inbred lines (RILs) derived from a cross ‘Rye Selection 111 (high GW) × Chinese Spring (low GW)’. The RILs and their two parental genotypes were evaluated for GW in six different environments (three locations × two years). Genotyping of RILs was carried out using 449 (30 SSRs, 299 AFLP and 120 SAMPL) polymorphic markers. Using the genotyping data of RILs, framework linkage maps were prepared for three chromosomes (1A, 2B, 7A), which were earlier identified by us to carry important/major genes for GW following monosomic analysis. QTL analysis for GW was conducted following genome-wide single marker regression analysis (SMA) and composite interval mapping (CIM) using molecular maps for the three chromosomes. Following SMA, 12 markers showed associations with GW, individual markers explaining 6.57% to 10.76% PV (phenotypic variation) for GW in individual environments. The high grain weight parent, Rye Selection111, which is an agronomically superior genotype, contributed favourable alleles for GW at six of the 12 marker loci identified through SMA. The CIM identified two stable and definitive QTLs, one each on chromosome arms 2BS and 7AS, which were also identified through SMA, and a third suggestive QTL on 1AS. These QTLs explained 9.06% to 19.85% PV for GW in different environments. The QTL for GW on 7AS is co-located with a QTL for heading date suggesting the occurrence of a QTL having a positive pleiotropic effect on the two traits. Some of the markers identified during the present study may prove useful for marker-assisted selection, while breeding for high GW in common wheat.     

Transgressive segregation for phenological traits in barley explained by two major QTL alleles with additivity

Quantitative trait loci (QTL) analysis can contribute to a deeper understanding of crop phenology. The parents of a barley mapping population have similar growth and development profiles, but the progeny show transgressive segregation for phenological traits. These phenotypes were measured in eight field experiments, using different planting dates over 3 years. Five QTL, on four chromosomes, were detected for anthesis date. The four maturity QTL were on the same chromosomes as the anthesis QTL. Five QTL for grain filling were detected on all chromosomes. Three QTL, on chromosomes 1H and 2H, were detected for photoperiod sensitivity. Both parents contributed higher value alleles for all traits, except for photoperiod sensitivity. QTL epistasis was not significant. Two QTL explained most of the phenotypic variation for anthesis and physiological maturity. Non‐parental combinations of alleles at these loci account for the phenotypic transgressive segregation. Candidate genes for these QTL effects are eps2S (2H) and denso (3H). QTL for other traits had smaller effects and were coincident with genes and/or QTL for the same traits reported in other germplasm.     

Simple sequence repeat markers reveal multiple loci governing grain-size variations in a japonica rice (Oryza sativa L.) mutant induced by cosmic radiation during space flight

Quantitative trait locus (QTL) for grain size traits that include grain length (GL), grain width (GW), grain thickness (GT) as well as thousand grain weight (TGW) were identified using F₂ population derived from a cross between a japonica cultivar Nongken58 and its large grain-sized mutant, ‘Dali’, which was selected in SP₂ generation of plants from Nongken58 seeds exposed to cosmic radiation upon space-flight, and then advanced it over eight successive generations by bagging the panicles to ensure self pollination. ‘Dali’ had similar GW and GT but 4.8 mm longer in GL, and 18.1 g heavier in TGW than those of Nongken58. Seven main-effect QTLs (M-QTLs) were identified for the grain size and weight traits. Among them, three M-QTLs, QGs3a and QGs3b for both GL and TGW, and QGs5 for GW, GT and TGW, which had strong additive effects on grain shape and grain weight, were validated in the two F₂ plant-derived F₃ populations. The three M-QTLs were found to be non-allelic to the cloned genes GS3, GL3.1, qSW5 and QGs5 by comparative mapping. However, there was only one pair of digenic epistasis involving QGs3b for TGW detected in this population. Interestingly, homozygous ‘Dali’ alleles at the QGs3a, QGs3b and QGs5 showed significant increase in the grain size and weight, suggesting these novel alleles of ‘Dali’ at the above three loci may be a very useful for marker-assisted improvement of grain quality for japonica cultivars.