Introduction
Identifying the driving forces promoting speciation is a key question in evolutionary biology (Mayr, 1942). Species pair with very low morphological differentiation but very significant genetic divergence (and vice versa ) offers excellent models for evolutionary biologists, which has improved our understanding of how species originate. Generally, limited gene flow is believed to hinder genetic divergence and lead to speciation (Rifkin, Castillo, Liao, & Rausher, 2019). However, researchers have focused on several pairs of closely related species diverged with gene flow (Aguirreliguori et al., 2019; Cooper, Whittall, Hodges, & Nordborg, 2010; Rifkin et al., 2019; Wu, Wang, Wang, Shrestha, & Liu, 2022; Zhao, Yin, Pan, & Gong, 2018; X. M. Zheng & Ge, 2010). Restricted gene flow fosters genetic and phenotypic divergence that may cause speciation (D. Schluter, 2001; Slatkin, 1987). The geographic isolation is generally considered the major driving force that impedes gene flow and promotes speciation, meanwhile, local adaptation can also generate barriers to gene flow and lead to ecological speciation (Aguirreliguori et al., 2019; Nosil, 2012; Rundle & Nosil, 2005; D Schluter & Conte, 2009). In addition, it is mysterious that some species pairs are morphologically similar while can be recognized using molecular markers (i.e. cryptic species)(Bickford et al., 2007). Comparing patterns of intra- and inter-specific genetic differentiation between morphological similarity species should be of great interest for understanding the process of their speciation.
Cycads, the most primitive living seed plants, include two families (Cycadaceae and Zamiaceae) with around 360 species of ten genera (Calonje, Stevenson, & Stanberg, 2022; Christenhusz et al., 2011). Although the cycad lineage is ancient, the extant cycad species have evolved recently with a synchronous global diversification during the late Miocene and are not older than 12 million years (Xiu Yan Feng, Liu, Chiang, & Gong, 2017; X. Y. Feng, Zheng, & Gong, 2016; Nagalingum et al., 2011; S. Y. Xiao, Ji, Liu, & Gong, 2020; Y. Zheng, Liu, & Gong, 2016). However, a recent study combining plastid phylogenomic, plate tectonic and fossil evidence indicated that the diversification of Cycadaceae (~69-43 million years ago, Palaeogene) was not that young (Jian Liu, Lindstrom, Marler, & Gong, 2022). The genusCycas L., the sole representative of the family Cycadaceae, occurs in the Australia, India, Malesian region, Japan and Southeast Asia, extending to Micronesia and Polynesia, Madagascar and East Africa (Jones, 2002). Biogeographic analyses strongly favor a South China origin for the living Cycas and an early dispersal to Indochina (L. Q. Xiao & Moller, 2015). However, controversial views emerged later. The ancestral area of Cycadaceae was recently inferred as a tropical region (i.e. Indochina) (Mankga, et al., 2020) or East Asia (Liu, et al., 2021). Cycas species have the presence of a crown of bird’s feather-like leaves, and their leaf bases all around the stem, which makes them look like palms in appearance. Two types of pinnae are presented in this genus, namely, pinnae simple and pinnae dichotomously divided. Most Cycas species are pinnae simple and only fourCycas species have dichotomously divided pinnae (Ken D. Hill, Nguyen, & Loc, 2004). Based on the key (Appendix S1) and morphological description (Ken D. Hill et al., 2004), C. bifida and C. micholitzii are the most two difficult sister species to distinguish in morphology (Figure S1), whereas their phylogenetic relationships are controversial (J. Liu et al., 2018; Y. Liu et al., 2022). According to the phylogeny of DNA fragment data, the two species had great genetic differentiation (J. Liu et al., 2018), while based on the transcriptome data, they showed sister species relationship (Y. Liu et al., 2022).
Cycas bifida is most closed with C. micholitzii but has larger male cones with microsporophyll tips that are generally without spines. Despite its original name (Appendix S1), the species has no affinity with C. rumphii . Geographically, C. bifida occurs in southwest China and the northern Vietnam, while C. micholitziioccurs in the Annam Highlands region of central Vietnam, southeast Laos and the northeast Cambodia (Ken D. Hill, 2008; Ken D. Hill et al., 2004). Their distribution areas are separated by the Red River Fault Zone (RRFZ) which is considered as a biogeographical barrier forCycas species distributed in this area (Jian Liu, Zhou, & Gong, 2015; Tang, 2004; Yang, Feng, & Gong, 2017; Y. Zheng et al., 2016). The RRFZ, defined by South China in the northeastern and Indochina in the southwestern, is proposed to be formed by India-Asia collision (Searle, 2006). For years, people have been debating it experienced left-lateral strike-slip in the first 30 - 16 Ma of collision (Tapponnier, Peltzer, Le Dain, Armijo, & Cobbold, 1982) with at least 15 km of displacement (Schoenbohm, Burchfiel, Liangzhong, & Jiyun, 2006) and presently undergone right-lateral motion since the late Neogene (8 - 7 Ma) with lower slip rate (Leloup et al., 1993; Schoenbohm, Burchfiel, & Liangzhong, 2006; Zuchiewicz & Cuong, 2009).
In the study reported here, we are aiming to find out the driving forces for these two species divergence by using multidisciplinary methods. We firstly used molecular approach to estimate the degree of species divergence among C. bifida and C. micholitzii . Then by integrating coalescent-based analysis (IMa), isolation by distance analysis (IBD), Barrier analysis, and Niche consistency detection, we tested the following specific hypotheses for how present patterns of genetic divergence arose: (i) they originated from the same ancestor or were one species before, and significant divergence between the twoCycas species was due to the long-term geographical isolation; (ii) they were originated from different ancestors. Their morphological similarities, especially the bifurcations of the pinnae, are a phenomenon of latter convergent evolution because of similar ecological niche.