Landscape features and geological history are hypothesized to strongly affect the biogeographical processes (Manel et al., 2003), and this can be especially true in freshwater systems (Leclerc et al., 2008). The river system evolution can shape and influence the distributions and phylogeographical pattern of freshwater-limited lineages among drainages (Schultheiß et al., 2014). The population structure and diversification of freshwater fishes are primarily influenced by the dynamics of river morphology, affecting connections between river basins and lakes. The separation of water bodies often leads to genetic divergence as a result of genetic drift and reproductive isolation (Husemann et al., 2012). And the intraspecific structure of riverine fish is usually clustered by river basins (Bartáková et al., 2015).

H. leucisclus often occur in high abundance in river and lake, because of high fertility, potentially reducing the amount of genetic drift through counteracting genetic differentiation among populations (Tibbets and Dowling, 1996). Further, passive dispersal by flooding may reduce genetic differentiation, at least for riverine fish within river basins (Van Leeuwen et al., 2013). Though the distribution of freshwater fish is often limited by water area, long-distance dispersal may occur when fish or their larvae are transported passively, either by accidentally or on purpose by humans. Hence, the broadly distributed H. leucisclus tends to show high connectivity and gene flow mediated by strong adaptation and fecundity, as well as their aggregative characteristic.

However, despite multiple modes of passive dispersal, strong population divergence is often observed among local populations in the freshwater system (Husemann et al., 2012), either resulting of natural geographic isolation or anthropogenic fragmentation. In China, the rapid uplift of Tibet Plateau changed the trajectory of the river system in China, and ultimately leading to the formation of a flowing eastward drainage system (Clark et al., 2004). The Yangtze River was built between Oligocene and Miocene, but the unification of the upper and middle Yangtze River in the Three Gorges mountain region has happened between Late Pliocene and Early Pleistocene (Zheng et al., 2013). However, it is generally believed that the drainage pattern of the modern Yellow River was closely followed by the great uplift of the Tibet plateau once again, about 1.7 million years ago (Mya), but the ultimate formation of the Yellow River was at about 0.15 Mya, when Sanmenxia was completely cut through by Yellow River and went through to east (Li et al., 2001). The development of major river systems in the Pearl River Basin was derived from the Yanshan tectonic movement, the connectivity between the Yangtze River and the Pearl River system was believed before the early Pleistocene (Miao et al., 2004). However, the Pearl River basins and Yangtze River basins are divided by the Yunnan-Guizhou Plateau and Nanling Mountains, which hold back the cold weather from the north, forming the warmer climate and well-watered tropical land in the Pearl River basins in China. As a result, the distinct and different evolutionary histories of the river systems can be expected to affect current patterns of diversity of the widely distributed freshwater species.

In this study, we used the partial mitochondrial cytochrome b gene (cytb, mitochondrial gene) and the recombination activating gene 2 (RAG2, nuclear DNA) and combined the data of previous studies to estimate the phylogeographic pattern and evolutionary history of H. leucisclus across the three major river systems in China, and to evaluate the conservation status and strategy of this species.

The role of past climate fluctuations and geological events, playing in extant species diversification and distribution, is one of the main focuses of phylogenetics and biogeography (Rull, 2011). One consequence of drastic climatic change and geological events is the frequent reconfiguration of landscape. Meanwhile, spatial and geographical barriers prevent gene flow, which in turn affects the phylogeographic pattern (Tittes and Kane, 2014). However, rivers can act as dispersal corridors for fish, promoting or counteracting geographical isolation and diversification (Vernesi et al., 2016).

Our results identified four distinct sublineages for H. leucisclus across different river systems in China with a significant phylogeographic patterns. One of the sublineages, sublineage C, had a very wide distribution across all the river systems in this study, even distributed in many isolated drainage systems, including Lancangjiang River, the upper Pearl River, the Fenhe River, and Weihe River (in Yellow River system), Huaihe River, and the river basins of southeastern China (Figure 1), suggesting it’s distributions did not completely match the river delimitation, as well as the geographic zone. The widely distributed sublineage C in different river systems could be attributed to their spatial proximity, which enables H. leucisculus populations from the Yangtze River basin to expand to nearby rivers easily by flooding in summer (Chen et al., 2017). However, the flooding cannot explain the strong admixture among far distant sampling locations across different river systems and even isolated drainage (Figure 2). Hence, anthropogenic translocations, especially the large-scale introduction of economically valuable fish half a century ago, offer alternatively possible explanations.

The formation of the Nanling Mountain Range could drive the separation of sublineage A from others (Wang L. H. et al., 2021), and the separation of sublineage B distributed in Sichuan Basin could be attributed to the isolation of populations by mountains barrier in the Southwestern and Central China (Figure 1). The Dabie Mountains-Eastern Qinling, Daba Mountains and Wushan Mountains have shaped the phylogeographic pattern of many other species, including Sinopotamon acutum (Fang et al., 2015), Microtus fortis (Gao et al., 2017), and Hyalessa maculaticollis (Liu et al., 2018). However, the populations from the Yellow River and Huaihe River basins had not yet formed independent lineage, although there are many watersheds existing in Qinling-Dabie Mountains. Furthermore, the populations from the middle and lower Pearl River basin and Hainan Island were unexpectedly found to have a close relationship. Considering the molecular data, the divergence time of H. leucisclus far postdated the formation of these mountains. For example, the ultimate formation of Nanling Mountain and the watershed between Yangtze River and Pearl River basin was dated to the Himalayan movement during the tertiary (Wang, 1993), suggesting that dispersal rather than vicariance has created the current distribution pattern of H. leucisclus in China.

Furthermore, the climate oscillations and geologic events during the quaternary could be a major driver of intraspecific divergence for H. leucisclus. Sublineage A with the highest nucleotide diversity (π = 0.00742, Supplementary Table 1) only existed in warm and moist tropical or subtropical areas, which provided important refugia for freshwater species during glaciations (Gao et al., 2012; Yan et al., 2013). Our results suggested that H. leucisclus originated in the Pearl River basin at 4.25–3.69 Mya (Figures 3A,B). And our data are consistent with the conclusion of previous studies (Chen et al., 2017). And the divergence time of lineage A was at approximately 2.57–2.27 Mya, which fit well with the separation of Upper Yangtze River and Pearl River during Late Pliocene and Early Pleistocene (Miao et al., 2004). It was speculated that the active river capture and reverse events may have caused the connection of the Yangtze River and the upper Pearl River in some regions before the Early Pleistocene. Because the upper Yangtze River originally drained into the South China Sea through the Paleo-Red River until it was captured by the middle and lower reaches of the Yangtze River and reversed its route to flow into the East China Sea (Perdices et al., 2005). The resulting river capture ultimately generated the isolation between the Yangtze River and Pearl River systems, and in turn, likely had a dramatic effect on the evolution of many freshwater species (Gu et al., 2019). Furthermore, the separation of sublineage B at approximately 1.42–1.62 Mya exactly coincided with the timescale of the formation of the current Yangtze River (Li et al., 2001). It was reported that the unification of the upper and middle Yangtze River in the Three Gorges Mountain region occurred between the Late Pliocene and Early Pleistocene (Zheng et al., 2013). Hence, the connection of the upper and middle Yangtze River facilitated species dispersal from west to east, similar to what was observed in the frog Quasipaa boulengeri (Yan et al., 2013) and the freshwater snail Bellamya (Gu et al., 2019). Therefore, the establishment of the eastward-flowing the Yangtze River and the changes in river systems could have provided opportunities for dispersal and gene flow between closely related freshwater species, especially fish. It was surprising that sublineage D had a relatively narrow and discontinuous distribution in Yangtze River and Yellow River basins. The divergence time estimation between sublineage C and D was at 0.96–0.83 Mya, basically matched with the Dagu Glaciation (1.1∼0.9 Mya) in China (Jing and Liu, 1999). The result of repeated episodes of range contraction and expansion during glacial cycling has been hypothesized to be important driving forces of vicariant speciation or intraspecific divergence in many fish species (Chen et al., 2017; Li et al., 2020).

The evolutionary significant units have been proposed for the assessments of biodiversity and have a significant impact on conservation program (Niemiller et al., 2012), especially the recognition of phylogeographic patterns has important implications for the accurate assessment of the conservation of genetic resources. According to the distribution of the four sublineages in different river systems, two sublineages were found in the Yellow River basin, three sublineages in the Yangtze River basin, and two sublineages in the Pearl River basin (Supplementary Figure 1), suggesting that the genetic structure of H. leucisclus has not been completely changed by recent human-mediated translocations and flooding. Based on our new findings, the conservation of H. leucisclus should be reassessed, as the current conservation strategies are performed under the assumption of only one species and one gene pool. Especially for the sublineage B in a couple of rivers in the upper Yangtze River basin with very low nucleotide diversity. In a previous study, the morphological diversification found in H. leucisculus along climate gradients is primarily caused by the divergent selection pressures among different environments (Wang L. H. et al., 2021). Given these indications and molecular information, we advocate different protection strategies for H. leucisculus in different river systems. There should be two protection units for H. leucisculus in the Yellow River basin, three protection units for H. leucisculus in the Yangtze River basin, and two protection units for H. leucisculus in the Pearl River basin. While only one protection unit for H. leucisculus is necessary in Huaihe River basin. Furthermore, a thorough survey of wild populations of these lineages, including an assessment of their ecology, will be required for a more precise evaluation of their conservation status.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

The animal study was reviewed and approved by the Animal Care Committee of Hunan Normal University.

QG conceived and designed the study. HZ and YS analyzed the data. HY, HZ, SL, and ZS carried out tissue dissection, DNA isolation, sample preparation for sequencing, and collect the mtDNA data from GenBank. QG and HZ drafted the manuscript. MW revised the language. All authors read and approved the final manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.