3.4 Molecular dating of cpDNA and demographic history
The BEAST-derived chronogram of 19 chlorotypes was supported by maximum PP, and revealed three diverging genetic lineages (Figure 3A), which correspond to the NW, SE and SW clades revealed by the chlorotype Median-Joining network (Figure 3B). The time-calibrated phylogenetic tree placed the split of P. heterotricha at c. 0.83 Myr (Figure 3A). The splitting time of the remaining chlorotypes from clades NW and SE was estimated at around 0.48 Myr, roughly at the end of early Pleistocene, and most other chlorotypes were dated between c.0.26-0.05 Myr (Figure 3). And two vicariance events were detected among three clades, first events gave rise to the genetic split in SW and the rest clades, second events offered opportunity to the genetic divergence in SE and NW clades (Figure S3).
The results of Tajimaʼs D test and Fuʼs Fs test were presented in Table 2 with the associated simulated P -values. The values for D was negative and Fs was positive for nrDNA sequences (D = -0.00585, P > 0.10; Fs= 0.538, P = 0.132), but all were positive for cpDNA sequences (D = 0.98381, P > 0.10; Fs = 9.119,P < 0.01) (Table 2), indicating an insignificant less and excess of rare nucleotide site variants for nrDNA and cpDNA sequences separately compared to the expectation under a neutral model of evolution, and an significant excess of rare haplotypes for nrDNA and cpDNA sequences over what would be excepted under neutrality. The Hierarchical mismatch analysis showed the distributions of differences for all populations, clade NW, SE and SW were ragged or multimodal (Figure S4), which rejected the hypothesis of demographic population expansion of these population clades.