**4. Discussion**

All eight HYVs used in this study were clustered into the *indica* group, based on *indica*–*japonica* SNP sets reported in a previous study [29]. Interestingly, some *japonica*-type SNPs were detected within the genomes of HYVs after resequencing analysis. Furthermore, collocations of 14 *japonica*-originated

genomic regions commonly present in Tongil-like HYVs, inherited from *indica*–*japonica* crosses, were investigated. This suggests that these *japonica*-type genomic segments were commonly and repeatedly selected during independent breeding programs in the temperate rice cultivation area. To investigate the role of these *japonica* segments in HYVs, a comparative study of the reported QTLs and representative gene selection were conducted. Eating quality, stress tolerance, and yield related traits might be main drivers for the selection of the rice HYV breeding program.

A total of 54 SNP markers, including 38 SNP markers developed from 38 selected genes for 14 common *japonica*-type regions and 16 trait-specific SNP markers, were used for the genotyping of 94 diverse varieties across *indica* and *japonica*. For the 38 SNP marker set, 16 SNP markers were located on chromosome 1, which was not identified as a *japonica* genomic segmen<sup>t</sup> in Takanari. Consequently, Takanari was clustered into the IND1 subgroup, although it showed a similar genotypic pattern to the other HYVs. The Chr1-1 and Chr1-2 regions were co-located with some abiotic tolerance and yield related QTLs. Furthermore, there were several selected genes that conferred abiotic stress tolerance and yield potential in the blocks. For instance, *Os01g0337100* (*OsTPS1*) reported an association with abiotic stress response and tolerance by knock-out and overexpression [40,41]. *Os01g0367100* (*PHD1*) was shown to be a gene that was involved in galactolipid biosynthesis and a ffected photosynthetic efficiency [42]. Recently, the e ffect of the haplotype of *PHD1* on grain yield was also reported using the 3K rice genome panel [18]. Takanari could not have acquired this *japonica* genomic segmen<sup>t</sup> from a di fferent natural environment and/or breeder's selection.

In addition, 18 varieties out of the 19 HYV-types, showed *japonica*-type *Wx<sup>b</sup>* allele on the SNP marker WAXY-TG, which was designed on the splicing site in the intron of *Wx* gene in the *japonica*-type region. The *Wx* gene only contains synonymous SNPs, although it is located within Chr6-1; thus, it was not selected in our study. We previously developed a functional SNP marker for the *Wx* gene [29]. The genomic region containing the *Wx* gene is a hotspot for grain quality [43] and has been selected during and after the domestication of rice [44]. The other genomic research using two Tongil-type varieties also showed a *japonica*-type SNP pattern on the common *japonica*-type region on chromosome 6 [45]. Tongil-type varieties showed medium amylose contents, approximately 19–20%, which is similar to that of non-waxy Korean *japonica* varieties [46,47]. Further, *Os06g0130400*, one of the selected genes in Chr6-1, was also reported as the gene controlling starch grain size in endosperm [48]. *Os06g0130000* and *Os06g0130100* were reported for resistance to rice blast and bacterial blight, and tolerance to drought and salt stress, respectively [49,50]. Therefore, Chr6-1, including the *Wx* gene, might be mainly selected for eating quality and latent stress tolerance.

When HYVs were developed by inter-subspecific hybridization, the breeders aimed at not only transferring some of the desirable characteristics, like resistance to lodging, blast, and yield, but also at retaining the ecological adaptability and eating quality of *japonicas* [51]. The *japonica* chromosomal introgression regions identified in this study were regarded as putative temperate region adaptable and improved the eating quality of *indica*. For this reason, the varieties developed by *indica*–*japonica* crosses could also be considered as 'temperate *indica*'. Recently, new elite rice varieties showing high yield potential and high grain quality were developed by the precise pyramiding of major genes controlling yield and grain quality traits [52]. Furthermore, there was an attempt to develop cold tolerant *indica* using an inter-subspecific cross and marker-assisted selection (MAS) [53]. In other words, breeding *indica* varieties, which are adaptable to the temperate region with high yield potential and good eating quality, can be e fficiently achieved through inter-subspecific crosses and marker-assisted selection using the SNP markers developed in this study. Nevertheless, to dissect the exact contribution of *japonica*-type regions in HYVs, a comprehensive genetic and physiological analysis, by applying the molecular markers developed in *japonica*-type regions to the segregation populations derived from cross between HYVs and *indica*, is necessary. In addition, the functional studies of genes in the regions, as well as the selected ones in this study, are also required.
