Genetic differentiation
AMOVA pointed out a high genetic differentiation level amongCerasus serrulata populations (cpDNA:F st=0.69325, ITS:F st=0.78692), signifying the existence of a phylogeographic structure (N ST=0.388 > G ST=0.328, P < 0.05). A similar high diversity level was also detected in C. dielsiana (ITS: F ST=0.7827) (Zhu et al. , 2019), but contrarily low levels were unraveled in C. pseudocerasus landraces (ISSR: F ST=0.308 ) (Chen et al., 2018), C. pseudocerasus containing both wild populations and landraces (cpDNA: F ST=0.24922, ITS:F ST=0.31571) (Chen et al., 2015), C. tomentosa (SSR: F ST=0.181) (He et al., 2015),C. lannesiana var. speciosa (cpDNA:F ST= 0.1655) (Kato et al. , 2011), C. mahaleb (RSPD: F ST= 0.1646) (Jordano and Godoy, 2000), C. campanulata (SSR: F ST=0.1477) (Lv, 2006; Su, 2007), C. jamasakura (SSR:F ST= 0.043) (Tsuda et al., 2009). This might because both C. serrulata and C. dielsiana are much more widely distributed than most of species in Cerasus . Additionally, adequate seed flow were also believed to weaken the interspecific divergence of C. pseudocerasus and other similar Cerasusspecies (Chen et al., 2015; Kato et al., 2011; Tsuda et al., 2009; Zhu et al., 2019).
Together of the cladograms and networks of both cpDNA and ITS sequences, finally two groups of C. serrulata within all the regions were defined: one mainly consisted of the northwest groups (NW), and the other was primarily composed of the left groups (SE+SW). Nevertheless, the northeast group (NE), involving population FHS in Liaoning Province and a foreign population K-NS in Korea, appeared a confusing phylogeographic pattern. It was assigned to group SE+SW in cpDNA analysis, but was related to group NW in ITS analysis. An earlier study based on nuclear SSRs, involving the main northeast and southeast populations of C. serrulata , indicated a consistent connection within the eastern regions involving population FHS and K-NS (Yi et al., 2018), which in accordance with the result of cpDNA analysis, making the distribution more puzzled. N ST/G ST analysis manifested no or weak phylogeographic structure within groups in northwest (NW) (N ST=0.305 <G ST=0.461, P > 0.05) and northeast (NE) (N ST=0.496 >G ST=0.473, P < 0.05). Actually,C. serrulata distributions were also recorded in Hebei Province and Heilongjiang Province in China, as well as in Japan, but we failed to obtain samples in these regions. Thus, the lack of the comprehensive sampling in Northeast China and Japan was mainly ought to be the reason of this confusion.