*2.5. Demographic (Bottleneck) Analyses*

Irrespective of the mutation model assumed (IAM, infinite allele model; SMM, stepwise mutation model; or TPM, two-phase mutation model), we found a significant excess of heterozygosity (*p* < 0.05) for one island (IP05) and one mainland population (EM04) in the TIL region, and one island population (ZI01) in the ZA region (Table 5), indicating that these populations may have experienced a recent bottleneck. However, a significant excess of heterozygosity was only found under the IAM model for one TIL population (IP08) and five ZA populations (ZI02, ZI07, ZM02, ZM03, ZM04; Table 5).

**Figure 3.** (**a**) The migration network among the 16 *Hedyotis chrysotricha* populations in the Thousand-Island Lake (TIL) region based on *G*ST (coefficient of gene differentiation) values and (**b**) significant relative migration relationships among populations (after 1000 bootstraps). Arrow thickness and color tone are proportional to the magnitude of the gene flow.

**Figure 4.** (**a**) The migration network among the 11 *Hedyotis chrysotricha* populations in the Zhoushan archipelago (ZA) region based on *G*ST (coefficient of gene differentiation) values and (**b**) significant relative migration relationships among populations (after 1000 bootstraps). Arrow thickness and color tone are proportional to the magnitude of the gene flow.


**Table 5.** Bottleneck analyses for a total of 27 populations of *Hedyotis chrysotricha*, including 16 from the Thousand-Island Lake (TIL) region and 11 from the Zhoushan Archipelago (ZA) region. The *p*-values are shown for the Wilcoxon signed rank tests, as evaluated under the infinite alleles model (IAM), the two-phase mutation model (TPM), and the stepwise mutation model (SMM), respectively.

Bold values indicate statistical significance (*p* < 0.05).
