Durum wheat (Triticum turgidum L. ssp. durum) is grown on about 8-10% of the world's wheat planted area, with the Mediterranean region and North America accounting for 60% of production. On the other hand, in Japan, durum wheat is rarely grown because the maturing period coincides with the short rainy season, which causes preharvest-sprouting and/or Fusarium head blight.

Therefore, it is essential to breed early-heading durum wheat varieties that can be harvested before the rainy season to expand the cultivation of durum wheat in Japan. Then, we tried to breed an extremely early heading durum wheat variety by introducing the non-functional alleles of PCL1 (pcl1-A1 and pcl1-B1) derived from a hexaploid wheat line ‘Chogokuwase’.

During the breeding process, some lines did not become extremely early heading, despite possessing both early flowering alleles of PCL1. This suggests the presence of novel heading time-related genes that interact with PCL1. In this study, genetic analysis was performed on several populations to determine the cause of the lack of extremely early-heading phenotypes in some lines.

We used three F2 populations (they were named as DF20, DF21, and DF22; 267 individuals in total) derived from crosses between extremely early-heading lines and non-extremely early-heading lines. The seed of populations was sown in the field of Okayama University on November 18, 2020. Genome-wide SNP information was obtained by Multiplexed ISSR genotyping by sequencing (MIG-seq; Nishimura et al., 2022). GWAS on heading date with 56,620 SNPs showed no SNPs with FDR<0.05.

Among these SNPs, we focused on the ones showing relatively higher peaks on chromosomes 3B and 5B. ANOVA using these peak SNPs showed that their genotypes significantly affected heading date. To perform BSA with higher density markers, 32 DF22 individuals (16 early- and 16 late-heading individuals) were subjected to degenerate oligonucleotide primer MIG-seq (dpMIG-seq), an improved method of MIG-seq (Nishimura et al., 2024).

Using 76,127 SNPs detected by dpMIG-seq, small peaks were observed on chromosomes 6B and 7A; markers near the 6B peak had a significant effect on heading date but this effect was small, while markers at the 7A peak had no significant effect on heading date.

Based on these results, it is expected that many genes with small effects, including QTLs on chromosomes 3B, 5B, and 6B, are segregating in the F2 populations in this study, and that their pyramiding effects are causing segregation of heading date.