3. Results
3.1.Genetic polymorphism and haplotype network
COI gene (658 bp) was successfully obtained from 346 individuals. The variable sites included 12 singleton variable sites and 33 parsimony information sites. 36 unique haplotypes were derived from 346 individuals. The distribution of different haplotypes based on COI was showed in Fig.2. The haplotype distributions were dividing three starry shapes in the network. According to the haplotype network, the geographical populations were further divided into 85 populations and this was done for further analysis of individuals from different lineages in the same collection site. 85 populations belonged to three groups, which included NE, SE and SN group. NE contained the individuals that produced with egg diapause and mainly lived in the northern regions. SE included the populations with egg diapause and mainly lived in the southern regions. SN group was consisted with individuals that produced with nymph diapause and mainly lived in the southern regions (Table 2). The location of three groups was showed in the Fig.1. There was no sharing haplotype among three groups. Hap24, Hap 2, Hap 4 and Hap 11 haplotypes were the most frequent haplotypes, characterizing 25.72%, 25.14%, 17.92%, 4.05% individuals, respectively. Hap 2 and Hap 11 were the ancestral haplotypes of NE group, and Hap11 was disjoint from Hap 2, which suggested was differentiated due to the long distance. Hap 4 and Hap 24 were the ancestral haplotypes of SE and SN groups, respectively. There were two lineages in populations of SDDY, but the haplotypes of the residents of SDDY belonged to the SE were the tips of Hap 2, thus the SE lineage may immigrate lately. In additional, the haplotypes from the United States were the tips of the gene tree, Hap 5 and Hap 8 were originated from the NE group. Populations of USAGA3 and USATN3 shared hap 5 with JPDB3, JPSH3 and JPSS3 in Japan. Hap 7 was from SE group, populations of USAMO1 and USAVA1 shared hap 7 with SDDY1, ZJTM1 and SHBS1 in China (Fig.2 and Table 2).
Pairwise FST among SE, NE and SN groups range from 0.87072 to 0.95084 (P < 0.001 Table 4), which suggested that the three groups were differentiated significantly. Isolations between SE, SN and NE group were consistent with results of AMOVA. The partitioning of total genetic variation in three clades using AMOVA indicated 93.23% diversity among groups, 2.78% within populations, and 4.00% among populations within groups (Table 5), suggesting thatV. micado has significant genetic differentiation at the different mode of life cycle and geographical separation of northern and southern regions. According to the zoogeographic regions, 85 populations were divided into 8 regions, including NEC, NC, SWC, CC, VK, KJ and USA. The partitioning of total genetic variation in nine clades using AMOVA indicated 24.94% diversity among groups, 69.56% among populations within groups, and 5.49% within populations (Table 6). The results were consistent with the network based on COI, which meant the different modes of life cycle accounted for the most of genetic variation.
Based on the CytB gene, 579 bp fragments were successfully obtained from 162 individuals. The variable sites included 18 singleton variable sites and 31 parsimony information sites. 36 unique haplotypes were derived from 162 individuals. The distribution of different haplotypes was showed in Fig.3. All individuals were divided into two groups significantly. Hap 12 and Hap 13 were the ancestral haplotypes of northern China (NC), and Hap 4 was the ancient haplotypes of southern China (SC). Coincidentally, all the individuals in the NC group were the members of the NE group defined by the COI, and others were in the SE and SN group. There were two lineages in SDDY and SHBS, which mainly including Hap1, Hap15 and Hap12. While Hap1 and Hap 15 were the tips of Hap 4, thus the NE lineage may be the native in SDDY and SHBS (Fig.3). This result was consisted with the assumption based on COI.
The partitioning of total genetic variation in different regions indicated 41.11% among populations within groups, 30.23% diversity among groups, and 28.66% within populations (Table 7). While the AMOVA result showed that the source of variation among groups categorized by three clades based on COI accounts 84.78% (Table 8).
3.2.Historical demographic changes
Based on COI, the negative of values of Tajima’ D suggested a relative constant population size. In SHBS3, SDDY3 and SDTA3, the current population size is stable. The NE lineage may be the native in SDDY and SHBS, which is consistent to the results of haplotype network (Table 2). Based on CytB, the negative and significant values of Tajima’ D indicated past population expansion in populations of GZXY (Guizhou province). Additionally, Hap 4 was the ancient haplotypes of populations of egg and nymph diapause. Individuals produced with nymph diapause in Zhejiang, Hainan and Guangxi (ZJTT, HNBS, HNJFL, HNWZ, HNCJ, HNXA, GXPM, GXSL and GXJX), whereas those were with egg and nymph diapause in Yunnan and Guizhou (YNKM and GZXY), which suggested the ancient haplotypes of SE group distributed there (Fig.3), and the positive Tajima’ D value based on COI and CytB suggested there may be a colonization of SE group in Guizhou (Table 2 and 9). Thus, it is possibility that SE group was originated from Yunnan and colonize to the other regions.
3.3.Divergence time estimation
The chronogram reconstructed with BEAST was based on COI (Fig. 4). The most recent common ancestor for all V. micado was dated at approximately 0.79 Myr (95% HPD: 0.46-1.13Myr). The diversification time between SN and SE clades was 0.50 Myr (95% HPD: 0.25-0.71Myr).