Tianshan Mountain provides a model for studying biological evolution and speciation. Here we assess the evolutionary history of the Ostrinia furnacalis and Ostrinia nubilalis, which are sympatric in the Yili River Valley in Xinjiang, China.

Our study is based on the historical gene flow analyses of two species by using three mitochondrial DNA (mtDNA, COI & COII & Cytb) and four nuclear DNA (nuDNA, EF-1α & Wingless & RPS5 & CAD) markers obtained from representatives of HC (Huocheng), YN (Yining), XY (Xinyuan) and MNS (Manasi).

Our results reveal that there is a strong asymmetrical gene flow pattern between the four populations. The population migratory pathways between these different populations show inflow into HC and YN, outflow from XY, and that MNS maintained a flow balance. Bayesian divergence time dating based on the COI gene suggest the genetic divergence between the two species in this area may have occurred in the late-Pleistocene (0.003–0.0127 Mya). Neutrality tests (Tajima’s D, Fu’s Fs) and mismatch distribution test results suggest that population expansion events may not have occurred in the recent past, which may follow the ‘mountain isolation’ hypothesis. The ML and BI trees of the mtDNA haplotype dataset show that ECB haplotypes are clustered together in a distinct clade and are clearly separate from ACB haplotypes. However, the geographical pattern of haplotype distribution is less clear and there is no strong correspondence between haplotypes and their geographical pattern for both ACB and ECB, implying that there has been frequent gene flow among the geographic populations in the Tianshan Mountains.

These findings confirm that geological factors play an important role in driving genetic patterns.

To evaluate the variation in genetic diversity among four geographical populations, we calculated the number of haplotypes (Nh), haplotype diversity (Hd), and nucleotide diversity (π) for each population using DnaSP v 6.12.03 (Rozas & Librado, 2017) based on the combined three mitochondrial genes dataset (MTD) and combined four nuclear genes dataset (NUD), respectively. Genetic distances within and between populations were calculated in MEGA v 11. 0. 8 (Tamura et al., 2021) employing the Kimura 2-parameter (K2P) model (Kimura, 1980; Kumar et al., 2016). In order to infer the historical gene flow and the direction of gene flow, we used Bayesian stochastic search variable selection (BSSVS) analyses (Su et al., 2015) based on the MTD.