To clarify the genetic basis of extremely early heading in rice, we conducted quantitative trait locus (QTL) analyses using F2 populations from two genetically wide cross combinations, Hayamasari/Kasalath (HaF2) and Hoshinoyume/Kasalath (HoF2). Hayamasari and Hoshinoyume are extremely early-heading japonica cultivars. Photoperiod sensitivity is completely lost in Hayamasari and weak in Hoshinoyume. Three QTLs, QTL(chr6), QTL(chr7), and QTL(chr8), for days-to-heading (DTH) in HaF2 were detected on chromosomes 6, 7, and 8, respectively, and QTL(chr6) and QTL(chr7) were detected in HoF2. On the basis of the chromosomal locations, QTL(chr6), QTL(chr7), and QTL(chr8) may be likely to be Hd1, Hd4, and Hd5, respectively, which had been detected previously as QTLs for DTH in an F2 population of NipponbarexKasalath. Alleles of QTL(chr7) decreased DTH dramatically in both Hayamasari and Hoshinoyume, suggesting that QTL(chr7) has a major role in determining extremely early heading. In addition, allele-specific interactions were detected between QTL(chr6), QTL(chr7) and QTL(chr8). This result suggests that not only allelic differences but also epistatic interactions contribute to extremely early heading. QTL(chr8) was detected in HaF2, but not in HoF2, suggesting that it determines the difference in DTH between Hayamasari and Hoshinoyume. A major QTL was also detected in the region of QTL(chr8) in QTL analysis using an F2 population of HayamasarixHoshinoyume. This result supports the idea that QTL(chr8) is a major factor that determines the difference in DTH between Hayamasari and Hoshinoyume, and is involved in photoperiod sensitivity.

Rice breeding programs in Hokkaido over the past 100 years have dramatically increased productivity and improved the eating quality of rice. Commercial varieties with high yield and good eating quality, such as Kirara 397, Hoshinoyume, and Nanatsuboshi, have been continuously registered since 1990. Furthermore, varieties with better eating quality using Wx1-1, which reduces amylose content to improve the taste of sticky rice, such as Oborozuki and Yumepirika, were registered in 2006 and 2008, respectively. However, to the best of our knowledge the genomic changes associated with these improvements have not been determined. Better understanding of the relationships between DNA sequences and agricultural traits could facilitate rice breeding programs in Hokkaido. Marker-assisted selection (MAS), which can select the plants with chromosomal regions tagged with DNA markers for desirable traits, is an advanced technology to manage genetic improvements. Here, we summarize the current states of MAS in rice breeding programs in Hokkaido before huge data sets of genome sequences using next-generation sequencing technology come into practical use in rice breeding programs.

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.

Plant breeding programs in local regions may have genetic and phenotypic variations that are desirable and shape adaptability during the establishment of local populations. Despite the characterization of genetic population structures in various kinds of populations, the effects of variations in phenotype on agro-economical traits currently remain unclear. In the present study, we evaluated phenotypic changes in 26 agro-economical traits among the local population during rice breeding programs in Hokkaido. Wide variations were observed in all 26 agro-economical traits with continuous distributions. In order to elucidate improvements in these agro-economic traits during rice breeding programs in Hokkaido, values were compared between genetic population structures. Traits were classified into four patterns based on the timing of significant differences. Patterns A and B showed significant differences once and twice, respectively. Pattern C gradually showed significant differences. Pattern D showed no significant differences for the desired directions. Based on the changes in phenotype observed in the present study and the genetic population structure for the local population in Hokkaido, a model of the artificial selection for phenotypes in genetic diversity among the local population during plant breeding programs has been proposed.

No abstract