III. A Clade Exclusive to Northeast Asia
Our analysis verifies the existence of a clade exclusive to Northeast Asia (Clade H), previously suggested by other studies (Fu et al., 2005; Hu et al., 2007; Borzée et al., 2017). Clade H was strongly supported (bootstrap value=100) and genetically distinct from other clades (Fig. 2 ). In previous genetic studies, B. gargarizans was treated as the B. gargarizans complex composed of several clades across a large area without differentiation according to region or altitude (Hu et al., 2007; Zhan & Fu, 2011; Borzée et al., 2017). These features made it difficult to understand the evolutionary history ofB. gargarizans .
The existence of the Clade H suggests new interpretations of the differentiation process of B. gargarizans in Northeast Asia. We estimate the divergence time estimate of Clade H to be 2.25 Ma (0.5 Ma–4.33 Ma; Fig. 5 ), in the Pliocene and early Pleistocene. During the early Pleistocene (Gelasian age), there were major geological events (landification due to fluctuations in sea level) (He et al., 2015) and dramatic climate change (glacial range expansion). Our multilocus haplotype network infers that the ancestor of Clade H was in southeastern China, as indicated by the mixed membership of Group 7 (SE and NE regions) (Fig. 3 ).
Previous studies suggested that faunal exchange between China and the Korean Peninsula occurred through the Yellow Sea land bridge at times of low sea levels (Zhang et al., 2016; Du et al., 2019), including B. gargarizans (Borzée et al., 2017). After dispersal into Northeast Asia, the subsequent rise of sea level and the expansion of glaciers would have isolated Clade H in a glacial refugium on the Korean Peninsula. A similar pattern of a glacial refugium on the Korean Peninsula was found in other organisms (Lee et al., 2008; Yoshikawa et al., 2008; Zhang et al., 2008; Kim et al., 2013; Borzée et al., 2017; Fong et al., 2020). Although we had limited sampling from North Korea, Heilongjiang Province (China), and the Russian Far East, there is a preliminary pattern indicating that South Korea is relatively diverse, which would support a scenario of range contraction into South Korea during a glacial cycle, followed by a range expansion northward during an interglacial cycle (Fig. 4 ). To verify this hypothesis, additional samples are needed from in the Northeast China, North Korea, and the Russian Far East.