1 INTRODUCTION

Marine fish are generally assumed to have high dispersal potential because of their high dispersal capability at both larval and adult stages, and the absence of physical barriers to dispersal (Caley et al., 1996; Hellberg, 2009). Theoretically, the higher dispersal capability of a species, the lower its inter-population genetic structure. Therefore, in marine fish species, especially migratory species, the signal of population differentiation is weak and difficult to detect because of high levels of gene flow (Gandra, Assis, Martins, & Abecasis, 2020). Understanding the distribution of genetic diversity of important commercial species is vital to implementing protection policies and management regulations (Araki & Schmid, 2010). Genetic diversity (within and among populations) greatly influences species adaptive potential to environmental changes, and ultimately determines their long-term resilience to ecological disturbances (Pauls, Nowak, Bálint, & Pfenninger, 2013). Additionally, knowledge of genetic structure is useful for understanding fish spawning migration routes, areas, and seasons. Such information can help fisheries managers define the spatiotemporal scales over which they can implement effective stock management and conservation plans (Bradbury, Laurel, Snelgrove, Bentzen, & Campana, 2008). Therefore, from a resource management perspective, understanding the patterns of genetic diversity and levels of gene flow among populations is a fundamental issue. The effectiveness of marine protected areas depends on both their ability to self-recruit (reproductive potential) and the spillover of adults and export of larvae to nearby fished areas (Harrison et al., 2012; Le Port et al., 2017). Over the last 30 years, genetic studies have become an essential tool for stock management and conservation of estuarine, coral reef, and coastal populations because genetic studies are useful for estimating genetic diversity and the ability to survive anthropogenic activities (e.g., overfishing, habitat degradation, eutrophication, invasive species, and pollution) (Gaither, Toonen, Robertson, Planes, & Bowen, 2010; Gill & Kemp, 2002; Machado, da Silva Cortinhas, Proietti, & Haimovici, 2020; Ruzzante, Taggart, & Cook, 1998; Ryman, Utter, & Laikre, 1995). Mitochondrial DNA (mtDNA) markers are widely employed to detect population structure in marine species because they have large number of copies, high mutation rates, generally maternal inheritance, and almost nonexistent recombination. The primary advantages of mtDNA are the inheritance pattern and nonexistent recombination: clonal inheritance through the maternal line allows tracing of speciation events over multiple generations, and male dispersal does not homogenize the population (Prugnolle & de Meeus, 2002). These factors make mtDNA markers particularly appropriate indicators of population genetic structure in marine organisms such as zooplankton and migratory fishes, which generally have high gene flow (Machado et al., 2020; L. Wang et al., 2013; Xu, Li, Wang, & Du, 2019).
Threadfin porgy, Evynnis cardinalis (Lacepède 1802), occurs in the Indo-West Pacific from Japan, Korea, and China to Vietnam and Indonesia (Z. Chen & Qiu, 2005). This species is primarily distributed from 30–60 m depth but also can occur to 100 m depth (Iwatsuki & Carpenter, 2014). This species is found over a wide range of bottom types, but is more common close to reefs or on rough bottoms. Small individuals are very common at some localities in shallow, sheltered bays; larger fish often occur in deeper water (Eggleston, 1974).Evynnis cardinalis is one of the main commercial fishing targets in bottom trawl fisheries in the northern South China Sea and is considered to have three geographical stocks in the northern South China Sea: the Taiwan Strait, South China Sea, and Beibu Gulf stocks.Evynnis cardinalis is a migratory fish that is captured year-round with significant seasonal differences among fishing areas. In Beibu Gulf, E. cardinalis undergoes seasonal migration and is found toward the northeastern shallow area of the gulf in late autumn and winter. Spawning takes place in northwest Weizhou Island in the spring, and the recruits disperse widely in the northern area of the gulf and begin to disperse south in late spring (Iwatsuki & Carpenter, 2014). However, this species exhibits life history characteristics, including late maturity and longevity, that predispose it to impacts from heavy exploitation. In the northern South China Sea, the Beibu Gulf stock of E. cardinalis declined 58% from 2001 to 2005, whereas the Taiwan Strait stock declined 62% from 1993 to 2008. Therefore, a recent IUCN Red List assessment reported E. cardinalis as endangered, and this species is mainly threatened by overexploitation (Iwatsuki & Carpenter, 2014). Previous studies on this species in Beibu Gulf have focused on feeding habits, growth and mortality, ecological distribution, phylogeny, and stock density (Cai, Chen, Xu, & Zhang, 2017; Z. Chen & Qiu, 2005; Z. Z. Chen & Qiu, 2003; D. Zhang, Shao, Su, & Jiang, 2007; K. Zhang et al., 2020; Y. Zhang, Dai, Yan, Yang, & Lu, 2014). However, no information is available to date on the genetic structure of E. cardinalis in the northern South China Sea, and this prevents reliable stock assessments and protection policy formulation. The objective of this study was to provide a population genetic analysis using a portion of the mitochondrial cytochrome c oxidase subunit I (COI) gene to assess the population genetic diversity pattern and historical demography, and estimate E. cardinalis expansion time in Beibu Gulf.