Heat tolerance (HT) of rice at anthesis is a key trait that ensures high and stable yields under heat stress. Finding the quantitative trait loci (QTLs) and gene loci controlling HT is crucial. We used relative spikelet fertility (RSF) as a measure of HT. The phenotypic values of RSF in 173 rice accessions were investigated in two environments and showed abundant variations. We performed a genome-wide association study on RSF using 1.2 million single nucleotide polymorphisms (SNPs). Five QTLs were significantly associated with RSF were identified, four were found in previously reported QTLs/genes, and one was novel. The novel QTL qRSF9.2 was mapped into the 22,059,984-22,259,984 bp region, which had 38 positional candidate genes. By combining the linkage disequilibrium analysis, the QTL region was narrowed to 22,110,508–22,187,677 bp, which contained 16 candidate genes. Among them, only gene LOC_Os09g38500 contained nonsynonymous SNPs that were significantly associated with RSF. In addition, accessions with large and small RSF values had corresponding respective high and low gene expression levels. Furthermore, the RSF of the CC allele was significantly higher than that of the TT allele. Hap 2 and Hap 3 can increase heat tolerance by 7.9 and 11.3%, respectively. Our results provide useful information that recommends further cloning of qRSF9.2 and breeding heat-tolerant rice varieties by marker-assisted selection.

Plant height (PH) and its component traits are critical determinants of lodging resistance and strongly influence yield in rice. The genetic architecture of PH and its component traits were mined in two mapping populations. In the natural population composed of 504 accessions, a total of forty simple sequence repeat (SSR) markers associated with PH and its component traits were detected across two environments via association mapping. Allele RM305-210 bp on chromosome 5 for PH had the largest phenotypic effect value (PEV) (−51.42 cm) with a reducing effect. Allele RM3533-220 bp on chromosome 9 for panicle length and allele RM264-120 bp on chromosome 8 for the length of upper first elongated internode (1IN) showed the highest positive PEV. Among the elongated internodes with negative effects being desirable, the allele RM348-130 bp showed the largest PEV (−7.48 cm) for the length of upper second elongated internode. In the chromosome segment substitution line population consisting of 53 lines, a total of nine QTLs were detected across two environments, with the phenotypic variance explained (PVE) ranging 10.07–28.42%. Among the detected QTLs, q1IN-7 explained the largest PVE (28.42%) for the 1IN, with an additive of 5.31 cm. The favorable allele RM257-125 bp on chromosome 9 for the 1IN increasing was detected in both populations. The favorable alleles provided here could be used to shape PH architecture against lodging.

Grain chalkiness directly affects the commercial value of rice. Genes related to chalkiness reported thus far have been discovered in mutants, but it has not been identified whether these genes can be used to improve rice quality by breeding. Therefore, discovering more quantitative trait loci (QTLs) or genes related to chalkiness in the rice germplasm is necessary. This study entails a genome-wide association study on the degree of endosperm chalkiness (DEC) and percentage of grains with chalkiness (PGWC) by combining 1.2 million single-nucleotide polymorphisms (SNPs) with the phenotypic data of 173 rice accessions. Thirteen QTLs for DEC and nine for PGWC were identified, of which four were detected simultaneously for both DEC and PGWC; further, qDEC11/qPGWC11 was identified as the major QTL. By combining linkage disequilibrium analysis and SNP information, LOC_Os11g10170 was identified as the candidate gene for DEC. There were significant differences among the haplotypes of LOC_Os11g10170, and the Hap 1 of LOC_Os11g10170 was observed to reduce the DEC by 6.19%. The qRT-PCR results showed that the gene expression levels in accessions with high DEC values were significantly higher than those in accessions with low DEC values during days 21–42 after flowering, with a maximum at 28 days. These results provide molecular markers and germplasm resources for genetic improvement of the chalkiness-related traits in rice.

Flag leaf angle (FLA) is an important outcrossing trait affecting the hybrid seed production in rice (Oryza sativa L.). Natural variation of FLA has been reported in rice, but the molecular basis for this variation is largely unknown. In this study, we investigated the phenotypic values of FLA in 353 rice natural accessions in six environments, which indicated that there was abundant phenotypic variation. We performed a genome-wide association study on FLA using 1.3 million single nucleotide polymorphisms (SNPs). A total of six quantitative trait loci (QTLs) were identified significantly associated with FLA, of which five were located in previously reported QTLs/genes and one was novel. We identified two causal gene loci for FLA, namely, OsFLA6 and OsFLA2; OsFLA6 was co-localized with the gene OsLIC. In addition, the accessions with large and small FLA values have corresponding high and low OsFLA6 expressions. OsFLA2^TT allele could increase significantly the seed setting percentage in hybrid F₁ seed production by field experiment. We also confirmed that the allele OsFLA2^TT increased the FLA compared with that of the isogenic line carrying allele OsFLA2^CC by transgenic complementation experiment. The allele frequencies of OsFLA6^GGand OsFLA2^TT decreased gradually with an increase in latitude in the Northern Hemisphere. Our results should facilitate the improvement of FLA of parents of hybrid rice.

The eating and cooking quality (ECQ) directly affects the taste of rice, being closely related to factors such as gelatinization temperature (GT), gel consistency (GC) and amylose content (AC). Mining the quantitative trait loci (QTLs), and gene loci controlling ECQ-related traits is vital. A genome-wide association study on ECQ-related traits was conducted, combining 1.2 million single nucleotide polymorphisms (SNPs) with the phenotypic data of 173 rice accessions. Two QTLs for GT, one for GC and five for AC were identified, of which two were found in previously reported genes, and six were newly found. There were 28 positional candidate genes in the region of qAC11. Based on a linkage disequilibrium (LD) analysis, three candidate genes were screened within the LD region associated with AC. There were significant differences between the haplotypes of LOC_Os11g10170, but no significant differences were found for the other two genes. The qRT-PCR results showed that the gene expression levels in the accessions with high ACs were significantly larger than those in the accessions with low ACs at 35d and 42d after flowering. Hap 2 and Hap 3 of LOC_Os11g10170 reduced the AC by 13.09% and 10.77%, respectively. These results provide a theoretical and material basis for improving the ECQ of rice.

Milling quality directly affects production efficiency in rice, which is closely related to the brown rice recovery (BRR), the milled rice recovery (MRR) and the head milled rice recovery (HMRR). The present study investigated these three traits in 173 germplasms in two environments, finding abundant phenotypic variation. Three QTLs for BRR, two for MRR, and three for HMRR were identified in a genome-wide association study, five of these were identified in previously reported QTLs and three were newly identified. By combining the linkage disequilibrium (LD) analyses, the candidate gene LOC_Os05g08350 was identified. It had two haplotypes with significant differences and Hap 2 increased the BRR by 4.40%. The results of the qRT-PCR showed that the expression of LOC_Os05g08350 in small-BRR accessions was significantly higher than that in large-BRR accessions at Stages 4–5 of young panicle development, reaching the maximum value at Stage 5. The increase in thickness of the spikelet hulls of the accession carrying LOC_Os05g08350^TT occurred due to an increase in the cell width and the cell numbers in cross-sections of spikelet hulls. These results help to further clarify the molecular genetic mechanism of milling-quality-related traits and provide genetic germplasm materials for high-quality breeding in rice.