COI clade and haplotype frequencies in the eight regions
The COI haplotypes were divided into two clades in the Bayesian phylogenetic tree (Figure 2). The demarcated vegetation maps, with clade and haplotype composition data, are shown in Figure 3a. Regions A, C, and D showed similar ratios of clade 1 to clade 2. Clade 2 was slightly more highly represented in regions B and F than in regions A, C, and D. Even though region E belongs to the Korean peninsula, all haplotypes from the region formed a group with region H (Japan) and region G (USA) haplotypes in clade 2. Moreover, the ratio of the L morph to the S morph in region E was similar to that in region H (Japan).
Hap 1 was a dominant haplotype in regions A, B, and C. In region D, the frequency of hap 1 was lower than that in the other regions (Figure 3b). In Korean populations, region E (Yeosu-si), in which the S morph can be found, and region F (Jeju), which is isolated from the mainland of the Korean peninsula, had haplotype compositions distant from those of regions A to D. The haplotype compositions and frequencies in region H (Japan) were different from those in Korea (Figure 3b).