Based on northern (NOL) and southern (SOL) mitochondrial lineages, recently, it proposed the new species Ammodytes heian and revived the species name Ammodytes japonicus to describe sand lances from the northwestern Pacific Ocean. This study used molecular methods to investigate genetic relationships between the two sand lance species in Korea and Japan. In total, 154 specimens were collected from four locations in Korea (Baengnyeongdo in the Yellow Sea, Tongyeong in the Korean Strait, and Jumunjin and Gijang in the East Sea), and 50 specimens were collected from a single location in Japan (Wakkanai in the Okhotsk Sea). Mitochondrial DNA analysis demonstrated that the individuals from Baengnyeongdo and Tongyeong all belonged to the SOL, whereas those from Gijang, Jumunjin, and Wakkanai included individuals from both the NOL and SOL (over 75% NOL). Population structure analyses were performed on the same individuals using seven microsatellite DNA markers. The population structure analysis based on 201 specimens identified two clusters (named as northern group and southern group), with the admixture proportion (q) of < 0.1 for the northern group in the Backyeongdo and Tongyeong sand lances and < 0.1 for the southern group in the Wakkanai sand lances. The high heterogeneity indicated that the former was probably A. japonicus and the latter probably A. heian. However, the admixture proportion in the Jumunjin and Gijang sand lances was 0.71–0.75 for the southern group, indicating that hybridization and unidirectional introgression from SOL to NOL occurs in southwestern margin of the East Sea. Our findings illustrate the speciation process based on different patterns of gene flow between Korean and Japanese sand lance, which is strongly influenced by both the paleo-climatic change and the contemporary local oceanic current pattern.
Citation: Kim J-K, Bae SE, Lee SJ, Yoon MG (2017) New insight into hybridization and unidirectional introgression between Ammodytes japonicus and Ammodytes heian (Trachiniformes, Ammodytidae). PLoS ONE 12(6): e0178001. https://doi.org/10.1371/journal.pone.0178001
Editor: Tzen-Yuh Chiang, National Cheng Kung University, TAIWAN
Received: February 3, 2017; Accepted: May 5, 2017; Published: June 5, 2017
Copyright: © 2017 Kim et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant files are available from the Figshare (http://figshare.com); DOIs 10.6084/m9.figshare.4797538 (Supporting information), 10.6084/m9.figshare.4797535 (Table 5), 10.6084/m9.figshare.4797532 (Table 4), 10.6084/m9.figshare.4797526 (Table 3), 10.6084/m9.figshare.4797520 (Table 2), 10.6084/m9.figshare.4797271 (Table 1), 10.6084/m9.figshare.4797268 (Fig.4), 10.6084/m9.figshare.4797256 (Fig.3), 10.6084/m9.figshare.4797253 (Fig.2), and 10.6084/m9.figshare.4796602 (Fig.1).
Funding: This research was supported by the Marine Fish Resources Bank of Korea (MFRBK) under the Minstry of Oceans and Fisheries, Korea (http://www.mof.go.kr) and also supported by the National Institute of Fisheries Science, Korea (R2017030). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Interspecific genetic exchange and hybridization are important factors in the evolution of some species [1, 2]. Many recent studies have investigated hybridization and introgression in marine fishes [3–5]. Hybridization may lead to introgression and speciation and can also reverse or accelerate speciation [2, 6–7]. Generally, patterns of introgression vary across genomes and depend on selection (e.g., unidirectional or bidirectional selection) and the parent species [3, 8]. Turner  suggested that, if introgression occurs frequently, then gene flow could eventually lead to loss of distinctiveness in populations and their development into geographical variants or subspecies.
Hybridization occurs more often when the ranges of closely related species overlap [3, 5], and asymmetric introgression may take place when one species exists at low density and in sympatry with a closely related taxon . Introgression by unidirectional gene flow can lead to a population losing its genetic identity, and bidirectional introgression can reverse speciation .
Cryptic species, which are morphologically identical but genetically distinct, may be generated allopatrically or sympatrically. However, genetic identification of cryptic species can be considerably more difficult when they have recently diverged by introgression [12, 13]. The northwestern Pacific sand lance is one species that could have resulted from introgressive hybridization or directional introgression because the possibility of hybridization between species of the taxonomically revised genus Ammodytes remains.
The taxonomic status of the northwestern Pacific sand lance was first questioned by Turanov and Kartavtsev , who suggested it was necessary to revise the scientific name Ammodytes personatus when referring to sand lances inhabiting the temperate zone of the Yellow Sea, the southern East Sea (Sea of Japan), and the waters south of Japan. Previous morphological and genetic studies on the Korean A. personatus suggested there were two lineages: (1) a northern lineage (NOL) and (2) a southern lineage (SOL) [15–17]. The number of vertebrae was a good taxonomic characteristic for distinguishing the two lineages: although the numbers partially overlap , 64–67 (modes = 65 and 66) in the NOL from the East Sea, and 62–65 (modes = 62, 63, and 64) in the SOL from the Yellow Sea and the Korean Strait. However, the NOL and SOL could not be regarded as separate species because of the presence of an admixture zone , spawning aggregation , and gene flow between the two lineages .
Nonetheless, the two lineages can be distinguished by their mitochondrial DNA (mtDNA) cytochrome oxidase I (COI) sequences, and Orr et al.  described them as separate species. That study proposed that the NOL from Fukushima and Wakkanai in Japan was a new species, Ammodytes heian, and Orr et al.  revived the species name Ammodytes japonicus to describe the SOL from the same area and southern Japan. Orr et al.  described the different morphological characteristics of the two species: the number of pored lateral-line scales (142–185 in A. heian vs. 132–166 in A. japonicus), the number of dorsal plicae (159–198 in A. heian vs. 144–182 in A. japonicus), body height (40–57% in A. heian vs. 32–53% in A. japonicus), upper jaw length (29–32% in A. heian vs. 30–39% in A. japonicus), and eye diameter (11–15% in A. heian vs. 12–20% in A. japonicus). However, these ranges overlap, making it difficult to distinguish between the two species using morphological features alone.
Many studies have used highly polymorphic microsatellite DNA markers to identify both individuals and populations [19, 20, 22]. Microsatellites are generally more variable than mtDNA and allozymes, therefore more useful for population genetic studies that include fingerprinting and genetic mapping [20, 23, 24]. In marine organisms, they can be used to analyze genetic diversity between wild and farmed organisms, verify parentage, and quantify population genetic variation in related and hybrid species [20, 25–27].
The aim of this study was to make comprehensive genetic comparisons that would reveal relationships among Korean and Japanese sand lances and to understand their microevolutionary history, which might include hybridization or unidirectional introgression.
Materials and method
This research did not involve endangered or protected species, and any locations, which performed in this research, are not required for specific permission. A total of 204 adult individuals were collected at four locations around Korean peninsula and one location in Japan during 2010–2015 (Fig 1). The sampling areas and dates are shown in Table 1. Muscle tissues were fixed and preserved in 95% ethanol. All Korean samples were deposited at the Ichthyology Laboratory in Pukyong National University (PKU), Korea.
Four locations in Korea: BA (Baengnyeongdo, n = 40), TO (Tongyeong, n = 32), GI (Gijang, n = 40), JU (Jumunjin, n = 40), and one location in Japan: JP (Wakkanai, n = 47). Open and solid circles indicate southern lineage (SOL) and northern lineage (NOL), respectively.
DNA extraction, PCR and sequencing
Genomic DNA was extracted from muscle tissue or pectoral fin clips using the AccuPrep Genomic DNA extraction Kit (Bioneer, Daejeon, Korea). Partial fragments of mitochondrial DNA cytochrome c oxidase subunit I (mtDNA COI) were amplified using VF2 and FishR2 (5′-TCW ACC AAC CAC AAA GAY ATY GGG AC-3′) primers . We carried out the PCR under the following conditions: initial denaturation for 5 min at 95°C, 35 cycles (denaturation for 1 min at 95°C, annealing for 1 min at 52°C, extension for 1 min at 72°C) and final extension for 5 min at 72°C. The amplified PCR product was electrophoresed in 1% agarose gel and visualized by an UV illuminator. PCR products were purified the using AccuPrep® PCR purification kit (Bioneer, Korea) and directly sequenced using the ABI BigDye terminator cycle Sequencing Ready Reaction Kit v.3.1 (Applied Biosystems Inc., USA). The mtDNA COI sequences were aligned using ClustalW  in BioEdit v.7 . The mtDNA COI Sequences were registered to the DDBJ / EMBL / GenBank nucleotide sequence databases (accession numbers KY773707–KY773908).
We designed seven microsatellite primers (Ape341, Ape327, Ape104, Ape308, Ape315, Ape349, Ape302) according to the methods of Ren et al. . The 5’-end of the forward primer for each primers was labeled with a fluorescent dye (6-FAM, HEX; Applied Biosystems, USA). PCR amplifications were carried out in 12.5μL reaction volumes contacting 50-100ng template DNA, 1.25 μL 10X Ex taq buffer, 2.5mM dNTP Mixture, 10μm of each primer, and 0.1μL Taq polymerase (TaKaRa, Shiga, Japan), and PCR conditions was performed according to Ren et al. . Successful amplification was assessed by electrophoresis using 1% agarose gel to separate DNA fragments. PCR products were diluted with deionized water, combined with a mixture of Hi-Di™ Formamide (Applied Biosystems) and internal size standard (GENESCAN 400HD ROX; Applied Biosystems) and denatured for 2 minutes at 95°C. Samples were run on an ABI 3730xL DNA sequencer and fluorescent peak data were sized and scores using Genemapper ver. 4.0 (Applied Biosystems, USA).
For mtDNA COI sequences, the number of haplotypes, haplotype frequency and pairwise FST were calculated using Arlequin v.126.96.36.199 . A minimum spanning network was constructed from the matrix of pairwise distances calculated among all haplotypes using Arlequin v.188.8.131.52, and TCS v.1.21 program. The phylogenetic tree was constructed with the neighbor-joining method and 1,000 bootstrap replications in MEGA 7. The sequences of A. japonicus (KJ137283 of Orr et al. ), A. heian (KJ137282 of Orr et al. ), A. personatus (KJ137281 of Orr et al. ), and A. hexapterus (KJ137280 of Orr et al. ) from the National Center for Biological Information database were used to confirm the taxonomic status of the Korean and Japanese sand lances.
For microsatellite analysis, in order to analyze genetic diversity for each locus, we calculated the number of allele (Na), observed heterozygosity (Ho), expected heterozygosity (He) and polymorphic information content (PIC) of the seven microsatellite DNA markers using Cervus v. 3.0.6  and calculated allelic richness (Ar) adjusted by population size and inbreeding coefficient (FIS) using FSTAT v. 184.108.40.206 (https://www2.unil.ch/popgen/softwares/fstat.htm). Hardy-Weinberg equilibrium within a population was verified using GENEPOP v. 4.3 [34, 35].
Genetic boundaries between populations were inferred using BARRIER software version 2.2 . For genetic population structure, FST values were calculated using Arlequin v.220.127.116.11. To investigate the pattern of population structure, the software STRUCTURE v.2.3.4  was used to identify clusters of genetically similar populations based on Bayesian approach. To infer most likely number of populations values of K (K = 1–7) were evaluated using 10 independent runs for each K value. Each replicate run consisted of 1,000,000 Markov Chain Monte Carlo (MCMC) repetitions; a burn-in 100,000 iterations to calculate inferred average values and standard deviations. To determine the optimum K value, ΔK was calculated using STRUCTURE HARVESTER . Discriminant analysis of principal components (DAPC) also provided posterior probabilities of population assignment for each individuals . To estimate the contemporary migration rates were estimated using Migrate-n version 3.6.11 . We set datatype to Microsatellite, migration prior (FST), constant mutation rate with a stepwise mutation, migration rate parameter (Theta and M to a maximum of 10), and a Bayesian analysis.
MtDNA sequences were used to categorize the 202 individual specimens collected from the five locations (Baengnyeongdo, Tongyeong, Gijang, and Jumunjin in Korea and Wakkanai in Japan). At Baengnyeongdo in the Yellow Sea and Tongyeong in the Korean Strait, only SOL specimens were collected, whereas specimens from both the SOL and NOL were collected at Gijang and Jumunjin in the East Sea and Wakkanai in the Okhotsk Sea, with NOL being dominant (> 75% of the total) (Fig 1). In total, 89 haplotypes were identified, forming three distinct clades were detected in the minimum spanning network analysis. The first clade (lineage A) consisted of individuals from all locations. The second clade (lineage B) consisted of individuals from three locations (Gijang, Jumunjin, and Wakkanai). The third clade (lineage C) consisted of only one haplotype, and all of these specimens were from Wakkanai (Fig 2). Compared to sequences of our Ammodytes species with those of Orr et al. , the phylogenetic tree showed that lineages A, B, and C corresponded to Ammodytes japonicus, A. heian, and A. hexapterus, respectively (S1 Fig).
The size of each circle is proportional to the haplotype frequency, and the branch lengths are proportional to the mutational steps. Lineages A, B and C correspond to Ammodytes japonicus, A. heian and A. hexapterus of Orr et al. (2015).
Seven microsatellite markers were used for genetic analysis of the 201 specimens. This analysis determined the number of alleles (Na), observed (Ho) and expected (He) heterozygosity, polymorphic information content (PIC), probability value estimates for deviations from Hardy—Weinberg equilibrium, and inbreeding coefficients (FIS) (Tables 2 and 3). Average number of alleles (Na) ranged from 22.8 (Ape 104) to 34.4 (Ape 349). Allele frequencies for each locus showed that Japanese sand lance has both the highest in Ape 304 (22.5%) and the lowest frequency in Ape 341 (8.09%) (S2 Fig). Heterozygosity was high with Ho values ranging from 0.564 (Ape 308) to 0.912 (Ape 104), and with He values varying from 0.885 (Ape 308) to 0.972 (Ape 349). PIC was high in most locations, but the Japan was the lowest values (0.870). Overall genetic variability was high, with allelic richness (Ar) per samples averages ranging from 20.4 (Japan) to 26.2 (Gijang). At population level, both Ho and He values was shown in the lowest Japan (Ho = 0.728, He = 0.885), but the highest in Jumunjin (Ho = 0.770, He = 0.961). In HWE test, high level of HW disequilibrium was observed in most of seven loci. Three of the seven loci were found to be out of HWE in Japanese population unlike Korean population having the departures from HWE in one locus, and this might be due to relatively small number of populations in Japan (Tables 2 and 3).
Values of pairwise FST were significant for comparisons between the Baengnyeongdo individuals and those in Gijang, Jumunjin, and Wakkanai, similarly comparisons between Wakkanai and all the Korean locations were significant (Table 4; S1 Table). BARRIER analysis demonstrated a genetic barrier between the Korean and Japanese sand lance, supported by FST values (FST > 0.05, Fig 3). According to the STRUCTURE analysis, the ΔK value was greatest at K = 2 (Fig 4A). These results indicate that the sand lances inhabiting Korea and Japan represent two distinct genetic units. The structure plots had two clusters with ratios that varied depending on location. The individuals from Wakkanai occurred primarily in cluster 2 (> 90%), whereas those from the four Korean locations were found predominantly in cluster 1. Specimens from Baengnyeongdo and Tongyeong consisted mostly of individuals from cluster 1 (> 90%), but only 70.9–75.0% of specimens from Gijang and Jumunjin were from cluster 1 (Fig 4B; Table 5). The first axis of the DAPC separated the Korean and Japanese sand lances into two spatially distinct populations (Fig 4C). This result was consistent with the STRUCTURE analysis. In addition, the UPGMA tree showed that sand lances were divided into two distinct groups (Korean group and Japanese group), and Korean group was separated into two subgroups (Baengnyeongdo + Tongyeong subgroup and Jumunjin + Gijang subgroup) (S3 Fig). Therefore, the sand lances were separated into Korean (Baengnyeongdo, Tongyeong, Gijang, and Jumunjin) and Japanese (Wakkanai) populations. As displayed in Table 6, the highest levels of gene flow between locations was observed from Tongyeong to Gijang (Nm = 71.00), and the lowest from Jumunjin to Tongyeong (Nm = 18.33). In addition, the level of gene flow between Jumunjin and Wakkanai had low values (Nm = 19.00). This indicated that asymmetrical gene flow between two group, and a minimal amount of gene flow between the Korean and Japanese sand lances, consistent with the BARRIER analysis.
The red line shows detected a barrier. Each population is indicated by abbreviation; Baengnyeongdo (BA), Tongyeong (TO), Gijang (GI), Jumunjin (JU) and Japan (JP), respectively.
(A) Estimation of population clustering level from seven microsatellite genotypes following Evanno et al. (2005) criteria obtained from STRUCTURE HARVESTER. (B) Population structure inferred by Bayesian clustering method using the STRUCTURE program. The different colors represent two different clusters (K = 2). Vertical lines separate each lineage of 5 locations. (C) Scatterplots of the discriminant analysis of principal components of the microsatellite data for 5 locations. Individual genotypes appear as dots. Each population is indicated by abbreviation; Baengnyeongdo (BA), Tongyeong (TO), Gijang (GI), Jumunjin (JU) and Japan (JP), respectively.
Historic divergence vs. contemporary gene flow
A minimum spanning network analysis based on mtDNA COI sequences identified three distinct lineages (A, B, and C) among the Korean and Japanese sand lances. The genetic distances between these lineages were 7.2–8.5% (A—B), 7.7–8.5% (A—C), and 5.4–6.2% (B—C), respectively. Compared with the study findings reported by Orr et al. , lineages A, B, and C correspond to Ammodytes japonicus, A. heian, and A. hexapterus, respectively. The mtDNA results demonstrate that A. japonicus is distributed throughout Japan, China, and Korea, but the distribution of A. heian is limited to Korea and Japan, and that of A. hexapterus is restricted to Japan. The distribution of A. japonicus and A. heian partially overlap in the southwestern margin of the East Sea and in Wakkanai, suggesting the possibility of hybridization between these two species. A more powerful analytical procedure using microsatellite DNA markers will be required to define these relationships. The microsatellite markers demonstrated that all populations had a high degree of genetic variability (He = 0.885–0.963), and the pairwise FST values were significantly different among populations (FST = 0.028, P < 0.01). The STRUCTURE analysis demonstrated that the sand lances were clustered into two groups, but the structure plot indicated that all individuals included mixed ratios from each cluster. The Wakkannai sand lance consisted of more than 90% cluster 2 (northern group), whereas the Baengnyeongdo and Tongyeong sand lances each consisted of more than 90% cluster 1 (southern group), suggesting there may be very little gene flow between the northern group (Wakkannai, Japan) and the southern group (Baengnyeongdo and Tongyeong, Korea). In addition, DAPC plots showed genetically two groups corresponding to Japanese and Korean sand lance and this is consistent with that of the BARRIER analysis. According to Palero et al. , the DAPC analysis in European spiny lobster (Palinurus Elephas) showed a clear split into the Atlantic and Mediterranean populations, suggesting the barriers between the Atlantic and Mediterranean transition. Therefore, our results indicated the presence of strong boundary between the Korean and Wakkanai sand lances. Additionally, Ren et al.  suggested that the northern group of sand lances is mostly distinct from the southern group except those in Kashima and Mutsu Bay and that difference in the allele frequencies between the two groups might indicate very limited gene flow between the two. Using cytochrome b and microsatellite analyses, Muths et al.  demonstrated significant genetic differentiation between Ephinephelus merra populations living in the western Indian Ocean, which suggested the presence of a cryptic lineage. In this study, both the mtDNA and microsatellite analyses indicated significant genetic variation among the sand lances.
In agreement with our results, previous studies using mtDNA COI (genetic distance, d = 8.1%) and control region analysis (d = 5.8%) have demonstrated that the Korean sand lance is highly differentiated at different locations [15, 19]. Indeed, movement of the two distinct mtDNA COI lineages was restricted by geological barriers. Various studies have suggested that the genetic structure of marine organisms living in the northwestern Pacific Ocean was affected by Pleistocene glaciation [15, 43, 44]. The lowered sea level during periods of glaciation probably isolated the East Sea (Sea of Japan) from the rest of the northwestern Pacific Ocean, perhaps resulting in historic divergence and a high level of local adaptation of the two lineages. Also, Kim et al.  suggested that both SOL and NOL larvae are limited distribution due to a physical barrier formed by the Nakdong River as well as because of the intersection between two different oceanic currents (the Eastern Korea Warm Current and North Korean Cold Current). However, spawning areas and timing for the two sand lance lineages overlap considerably in the southwestern margin of the East Sea, where NOL and SOL larvae ratios are similar, and this indicated possibility of gene flow between the two lineages. In this study, most of loci showed high level of Hardy Weinberg disequilibrium, and especially, two loci (Ape302 and Ape 315) showed significant deviations from HWE in all populations. Ren et al.  found the deviations from HWE in some loci, and suggested that these could be affected by null alleles. In addition, Ren et al.  found the presence of null allele in locus or the dropout of large allele in the 12 deviation from HWE populations. In this study, the frequency of null allele ranged from 0.00 to 0.26, indicating the deviation from HWE could be associated with presence of null alleles. However, Ren et al.  insisted that these ms markers are useful for population structure study of sand lance despite presence of null alleles. Furthermore, recent study revealed that null alleles could not change the overall result of population structure using ms loci . The HW disequilibrium due to a deficiency of heterozygotes could be resulted by various factors such as inbreeding, selection, genetic drift or Wahlund effect [46–48]. The extent of HWE depends on migration pattern, selection, and isolation between two groups . The population structure of Killifish (Fundulus heteroclitus) using ms markers showed two distinct groups, which were consistent with previous study (mtDNA and allozyme loci), and the distribution and population structure between two groups were associated with differences of allele frequencies by fixation of clinal alleles . In case of herring in the Baltic Sea, clinal patterns of genetic differentiation between Baltic and Atlantic region occurred, and this pattern was related with higher or lower frequencies in certain alleles. Similarly, in this study, Japanese (Wakkanai) and Korean (Baengnyeongdo and Tongyeong) sand lance have well-defined groups in both mtDNA and ms loci, and this could be interpreted as two genetically heterogenous groups by allele fixation. According to Mallet , heterozygote deficiencies and linkage disequilibrium could be indicate that multiple distinct species have been pooled within a population. In case of brown trout Salmo trutta, putative hybrid populations have higher HW disequilibrium than pure ones, and these disequilibrium could be evidence of highly subdivided genetic structure . In addition, Bossu et al.  suggested that if there is a random mixing between genetically distinct two darter species (Etheostoma bison and E. caeruleum), linkage disequilibrium is expected to be high, and the more frequencies of hybrids have tendency to the more significant deviation from HWE. In this study, although two groups of sand lance have little gene flow, there are evidence of gene flow between two groups in the southwestern margin of the East Sea, indicating that genetically effective migration may occur. Especially, Gijang and Jumunjin sand lances have higher frequencies of mixing through gene exchange by migration, and this could affect their allele frequency and deviations from HWE. Therefore, this departures from HWE could be related to the several factors including the migration. In addition, the microsatellite analyses in this study that indicated the presence of gene flow also suggest that sand lances from the southwestern margin of the East Sea (Jumunjin and Gijang) have a less well-defined genetic structure. This may be explained by a variety of evolutionary mechanisms, including incomplete sorting of ancestral polymorphisms, and sex-biased hybridization . Consistent with our study, Han et al.  reported that secondary contact may occur between the two sand lance lineages in the waters surrounding Hokkaido and the Iwate Prefecture of Japan, where the Oyashio Cold Current and the Kuroshio Warm Current meet and the fish have poorly defined genetic structure due to the flow of currents. Ren et al.  also suggested that the oceanic currents may provide the opportunity for connectivity of among populations. Therefore, oceanic current fluctuations can have a significant influence on genetic exchanges between populations, and oceanographic processes may affect both adult migration and larval dispersal. Topographical factors are one possible explanation. The Korean Strait that separates Korea from Japan is very wide (175 km), whereas the Tsugaru Strait between Honshu and Hokkaido is relatively narrow (53 km), and gene flow between the two lineages across the Korean Strait was probably more common during the glaciation period. Therefore, the current population structure may be explained by both historical recolonization and current gene flow resulting from larval dispersal by oceanic currents.
Our results demonstrated that 71–75% of the Gijang and Jumunjin sand lances (mainly belong to NOL in mtDNA results) were from cluster 1 (southern group). The migration rate inferred from microsatellite markers was highest between Tongyeong (all comprise SOL) and Gijang (mostly comprise NOL), accordingly, unidirectional introgression occurs predominantly from the SOL (Yellow Sea, Korean Strait) to the NOL (East Sea). Han et al.  discovered admixture zones between the two lineages in the waters surrounding Hokkaido and the northeastern coast of Honshu in Japan, suggesting that secondary contact did occur. The admixture proportions (q) in Kashima (Iwate Prefecture) and Mutsu Bay were 0.226 and 0.105, respectively, suggesting that reproductive isolation between the two groups was not complete . In Bayesian admixture plot, most of the Icelandic eels showed on-going hybridization between American and European eels unlike having only a typical American haplotype in mtDNA, and this might be due to asymmetric introgression followed by subsequent backcrossing . Similarly, most individuals of Merluccius capensis and M. paradoxus from North Benguela have mixed genotypes equivalent to 0.36 and 0.2, respectively. In particular, all M. paradoxus individuals had M. capensis COI sequences, and this resulted from recurrent backcrosses between M. capensis × M. paradoxus hybrid females and M. paradoxus . Therefore, backcrossing between descendants of A. japonicus × A. heian hybrid females and A. japonicus males could have occurred in Gijang and Jumunjin.
Hybridization occurs frequently when spawning areas or periods overlap [3, 5, 57, 58]. Bae et al.  describe how sea bass (Lateolabrax japonicus) and spotted sea bass (Lateolabrax maculatus) have overlapping distributions in the Korean Strait and how some intermediate individuals (Lateolabrax sp.) of mixed genetic composition were formed by unidirectional introgression. Additionally, two North American Atlantic hake species, Merluccius bilinears and M. albidus, demonstrated a high level of introgressive hybridization where their ranges overlapped, suggesting that long-term hybridization and backcrossing had occurred . Kim et al.  suggested that the spawning areas of the two lineages of Korean sand lance overlap considerably in the southern East Sea, and the unidirectional hybridization between these two groups is feasible.
Hybridization has great potential for rapidly introducing variability into recipient populations if the barriers to recombination can be overcome . For example, Serranus cabrilla individuals in Cabo de Gata (CG) and Cabo de Palos (CP) from the Atlantic group were comparable to 21.7% of fish with an admixed genotype from a Mediterranean population. This may have occurred when migrants were carried south by oceanic currents. Therefore, genetic population structures may be significantly influenced by present current regimes that traverse oceanographic barriers . Strong directional gene flow may occur where there is dispersal in one direction due to selective barriers. Kim et al.  suggested that SOL larvae may be readily dispersed from the Yellow Sea to the Korean Strait by the Western Korea Cold Waters, while NOL larvae may not be transported from the East Sea southwest to the Korean Strait against the Eastern Korea Warm Current prevailing in the surface layer; indeed, this could facilitate unidirectional introgression between the two genotypes despite the weak biogeographical barrier between the Yellow Sea / Korean Strait (Baengnyeongdo and Tongyeong) and East Sea (Jumunjin and Gijang). Therefore, the present genetic structure of marine organism populations may be heavily influenced by local oceanic currents. The seasonal and directional changes of a variety of fronts and currents may be extremely important in determining gene flow patterns and population structure [59, 60].
Hybridization can also occur between species when one species is rare and individuals are forced to find mates from a closely related species [2, 61]. Kim et al.  reported that the number of sand lance larvae decreased significantly in the southern East Sea. Therefore, it is possible that interspecific hybridization occurs in preference to assortative mating when there are few individuals from a particular group. Interspecific hybridization driven by low numbers of individuals might be common throughout the hybrid zone in the southwestern margin of the East Sea (Gijang and Jumunjin).
Another cause of asymmetric gene flow is sex-biased dispersal or philopatry. Different genetic patterns in maternally inherited markers (e.g., mtDNA) and bi-parentally inherited markers (e.g., microsatellites) may arise because of sexual differences in spawning behavior or simply by genetic drift and population bottlenecks . If female-biased dispersal occurs, mtDNA will be homogeneous, but nuclear DNA (nDNA) will exhibit differentiation. In contrast, if male-biased dispersal occurs, mtDNA will differentiate, but nDNA will not . Therefore, further studies will be required to determine whether sex-biased dispersal has a role to play in these relationships.
Orr et al.  identified two mitochondrial lineages (the NOL and SOL) in sand lances from Fukushima (Iwate Prefecture) and Wakkanai (Hokkaido). They proposed that the NOL corresponded to the new species Ammodytes heian and revived the species name Ammodytes japonicus to describe the SOL. Orr et al.  reported that these two species could be morphologically distinguished by the number of pored lateral-line scales, the number of dermal plicae, body height, upper jaw, and eye diameter. However, these characteristics were overlapped. Baltic herring differ regionally in their weight, length, number of vertebrae, and number of pectoral fin rays, and these variations may reflect phenotypic plasticity due to genetic differentiation in addition to differences in environmental conditions during their development . Therefore, morphological variations between southern and northern groups of sand lance may certainly be influenced by environmental factors (e.g., temperature) in addition to genetic ones.
The two lineages were regarded as separate species , probably because the genotypes of Japanese sand lances were mostly consistent with haplotypes from the all same areas except two areas (Kashima, Mutsu bay) . However, it is difficult to conclude that the two lineages of Korean sand lances are distinct species because gene flow occurs between the two in Gijang and Jumunjin. Kai et al.  performed amplified fragment length polymorphism marker analysis on three lineages (named as SOJ1, SOJ2 and SOJ3) of Careproctus rastrinus living in the northwestern Pacific Ocean and demonstrated that SOJ1 (group 1 in Sea of Japan) and SOJ2 (group 2 in Sea of Japan) were not separated by principal component analysis, suggesting that gene flow may have occurred by secondary contact. The study concluded that the two lineages (SOJ1 and SOJ2) were not cryptic species because, although their mtDNA was well differentiated, they could not be separated by nDNA analysis. Therefore, it is unlikely that the two sand lance lineages in the southwestern margin of the East Sea are cryptic species because they could not be differentiated by microsatellite analysis (Table 5, Fig 4). Instead, by being exposed to secondary contact before reaching complete reproductive isolation, the two sand lance lineages in the southwestern margin of the East Sea may be evolving toward genetic homogeneity.
Speciation generally occurs as a gradual process, by reproductive isolation and genetic differentiation between populations . Hybridization and introgression occur when reproductive isolation is incomplete. Although the NOL and SOL have different morphotypes , ontogenetic characteristics , and hatching types , it cannot be concluded that these two lineages in the southwestern margin of the East Sea have been completely separated by reproductive isolation. Based on this study and a comprehensive review of many previous ones, we believe the sand lance in the Yellow Sea and Korean Strait is A. japonicus, whereas the sand lance in the southwestern margin of the East Sea may be derived from the backcrosses between descendant of A. japonicus × A. heian hybrid females and A. japonicus males because of the significantly higher proportion of admixture (0.71–0.75). The present study serves as a follow-up study which confirms the hypotheses of Han et al. , Ren et al.  and Kim et al.  where the authors suggested incomplete reproductive isolation between the two sand lance species. Our study revealed that characteristics of local oceanic current may cause different pattern of gene flows even within the conspecific.
S1 Table. Pairwise Fst values between locations_lineages (below diagonal) and pairwise P values (above diagonal) in seven microsatellite loci.
S1 Fig. The phylogenetic tree based on mtDNA COI in 444 bp.
The Neighbor-joining tree was constructed under the K2P model, and the numbers on branches indicate bootstrap probabilities from 1,000 bootstrap replications.
S2 Fig. Allele frequency distributions for five locations of sand lances based on seven microsatellite loci.
Each population is indicated by abbreviation; Baengnyeongdo (BA), Tongyeong (TO), Gijang (GI), Jumunjin (JU) and Japan (JP), respectively.
We would like to thank Dr. John R. Hyde (NOAA), Dr. Mi Sook Hwang (NIFS) for their useful comments on this manuscript. We thank two anonymous reviewers for their valuable comments that improved the quality of this article.
- Conceptualization: JKK SEB.
- Funding acquisition: JKK SJL.
- Investigation: JKK SEB SJL MGY.
- Resources: JKK SJL.
- Supervision: JKK.
- Visualization: JKK SEB.
- Writing – original draft: JKK SEB SJL MGY.
- Writing – review & editing: JKK SEB.
- 1. Martinsen GD, Whitham TG, Turek RJ, Keim P. Hybrid populations selectively filter gene introgression between species. Evolution. 2001; 55: 1325–1335. pmid:11525457
- 2. Abbott R, Albach D, Ansell S, Arntzen JW, Baird SJ, Bierne N, et al. Hybridization and speciation. J Evol Biol. 2013; 26: 229–246. pmid:23323997
- 3. Yaakub SM, Bellwood DR, van Herwerden L, Walsh FM. Hybridization in coral reef fishes: introgression and bi-directional gene exchange in Thalassoma (family Labridae). Mol Phylogenet Evol. 2006; 40: 84–100. pmid:16581267
- 4. Von der Heyden S, Connell A. Evidence of hybridisation within the genus Chrysoblephus and conserved nuclear sequences of South African sparids (Teleostei: Sparidae). Afr J Marine Sci. 2012; 34: 505–510.
- 5. Bae SE, Kim J-K, Kim JH. Evidence of incomplete lineage sorting or restricted secondary contact in Lateolabrax japonicus complex (Actinopterygii: Moronidae) based on morphological and molecular traits. Biochem Syst Ecol. 2016; 66: 98–108.
- 6. Baack EJ, Rieseberg LH. A genomic view of introgression and hybrid speciation. Curr Opin Genet Dev. 2007; 17: 513–518. pmid:17933508
- 7. Canestrelli D, Bisconti R, Nascetti G. Extensive unidirectional introgression between two salamander lineages of ancient divergence and its evolutionary implications. Sci Rep. 2014; 4: 6516. pmid:25269625
- 8. Bernal MA, Gaither MR, Simison WB, Rocha LA. Introgression and selection shaped the evolutionary history of sympatric sister-species of coral reef fishes (genus: Haemulon). Mol Ecol. 2017; 26: 639–652. pmid:27873385
- 9. Turner GF. What is a fish species? Rev Fish Biol Fish. 1999; 9: 281–297.
- 10. Choleva L, Musilova Z, Kohoutova-Sediva A, Paces J, Rab P, Janko K. Distinguishing between incomplete lineage sorting and genomic introgressions: complete fixation of allospecific mitochondrial DNA in a sexually reproducing fish (Cobitis; Teleostei), despite clonal reproduction of hybrids. PLoS One. 2014; 9: e80641. pmid:24971792
- 11. Seehausen O. Conservation: losing biodiversity by reverse speciation. Curr Biol. 2006; 16: R334–R337. pmid:16682344
- 12. Forsman ZH, Barshis DJ, Hunter CL, Toonen RJ. Shape-shifting corals: molecular markers show morphology is evolutionarily plastic in Porites. BMC Evol Biol. 2009; 9: 45. pmid:19239678
- 13. Ross KG, Gotzek D, Ascunce MS, Shoemaker DD. Species delimitation: a case study in a problematic ant taxon. Syst Biol. 2010; 59: 162–184. pmid:20525628
- 14. Turanov S, Kartavtsev YP. The taxonomic composition and distribution of sand lances from the genus Ammodytes (Perciformes: Ammodytidae) in the North Pacific. Russ J Mar Biol. 2014; 40: 447–454.
- 15. Kim JK, Kim YS. Genetic diversity, relationships and demographic history of three geographic populations of Ammodytes personatus (Ammodytidae) from Korea Inferred from mitochondrial DNA control region and 16S rRNA sequence data. Genes Genom. 2006; 28: 343–351.
- 16. Kim JK, Park JH, Kim YS, Kim YH, Hwang HJ, Hwang SJ, et al. Geographic variations in Pacific sand eels Ammodytes personatus (Ammodytidae) from Korea and Japan using multivariate morphometric analysis. J Ichthyol. 2008; 48: 904–910.
- 17. Kim JK, Watson W, Hyde J, Lo N, Kim JY, Kim S, et al. Molecular identification of Ammodytes (PISCES, Ammodytidae) larvae, with ontogenetic evidence on separating populations. Genes Genom. 2010; 32: 437–445.
- 18. Han Z, Yanagimoto T, Zhang Y, Gao T. Phylogeography study of Ammodytes personatus in Northwestern Pacific: Pleistocene isolation, temperature and current conducted secondary contact. PLoS One. 2012; 7: e37425. pmid:22685527
- 19. Kim J-K, Lee SJ, Lee W-C, Kim JB, Kim HC. Restricted separation of the spawning areas of the two lineages of sand lance, Ammodytes personatus, in the Yellow and East Seas and taxonomic implications. Biochem Syst Ecol. 2015; 61: 319–328.
- 20. Ren GJ, Hu JJ, Gao TX, Han ZQ. Population structure and genetic diversity of Ammodytes personatus in the Northwestern Pacific revealed by microsatellites markers. Biochem Syst Ecol. 2015; 61: 303–311.
- 21. Orr JW, Wildes S, Kai Y, Raring N, Nakabo T, Katugin O, et al. Systematics of North Pacific sand lances of the genus Ammodytes based on molecular and morphological evidence, with the description of a new species from Japan. Fish Bull. 2015; 113: 129–156.
- 22. Xue D, Wang H, Zhang T, Liu J. Population genetic structure and demographic history of Atrina pectinata based on mitochondrial DNA and microsatellite markers. PLoS One. 2014; 9: e95436. pmid:24789175
- 23. Banks MA, Rashbrook VK, Calavetta MJ, Dean CA, Hedgecock D. Analysis of microsatellite DNA resolves genetic structure and diversity of chinook salmon (Oncorhynchus tshawytscha) in California s Central Valley. Can J Fish Aquat Sci. 2000; 57: 915–927.
- 24. Rabone M, Lavery S, Little A, Clements K. Discordance between nuclear and mitochondrial DNA analyses of population structure in closely related triplefin fishes (Forsterygion lapillum and F. capito, F. tripterygiidae) supports speciation with gene flow. Mar Biol. 2015; 162: 1611–1624.
- 25. Mojekwu T, Anumudu C. Microsatellite markers in Aquaculture: Application in Fish population genetics. IOSR J Environ Sci Toxicol Food Technol. 2013; 5: 43–48.
- 26. Putman AI, Carbone I. Challenges in analysis and interpretation of microsatellite data for population genetic studies. Ecol Evol. 2014; 4: 4399–4428. pmid:25540699
- 27. Magalhaes IS, Ornelas-Garcia CP, Leal-Cardin M, Ramirez T, Barluenga M. Untangling the evolutionary history of a highly polymorphic species: introgressive hybridization and high genetic structure in the desert cichlid fish Herichtys minckleyi. Mol Ecol. 2015; 24: 4505–4520. pmid:26175313
- 28. Ivanova NV, Zemlak TS, Hanner RH, Hebert PD. Universal primer cocktails for fish DNA barcoding. Mol Ecol Notes. 2007; 7: 544–548.
- 29. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994; 22: 4673–4680. pmid:7984417
- 30. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. 1999; 41: 95–98.
- 31. Ren G, Hu J, Bao Z, Jiang X, Gao T. Isolation and characterization of eleven polymorphic microsatellite markers of sand lance (Ammodytes personatus). Conserv Genet. 2009; 10: 1837.
- 32. Excoffier L, Laval G, Schneider S. Arlequin (version 3.0): an integrated software package for population genetics data analysis. Evol Bioinform. 2005; 1: 47–50.
- 33. Kalinowski ST, Taper ML, Marshall TC. Revising how the computer program CERVUS accommodates genotyping error increases success in paternity assignment. Mol Ecol. 2007; 16: 1099–1106. pmid:17305863
- 34. Raymond M, Rousset F. GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. J Hered. 1995; 86: 248–249.
- 35. Rousset F. genepop'007: a complete re-implementation of the genepop software for Windows and Linux. Mol Ecol Resour. 2008; 8: 103–106. pmid:21585727
- 36. Manni F, Guérard E, Heyer E. Geographic patterns of (genetic, morphologic, linguistic) variation: how barriers can be detected by using Monmonier's algorithm. Hum Biol. 2004; 76: 173–190. pmid:15359530
- 37. Evanno G, Regnaut S, Goudet J. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol. Ecol. 2005; 14: 2611–2620. pmid:15969739
- 38. Earl DA. STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method. Conserv Genet Resour. 2012; 4: 359–361.
- 39. Jombart T, Devillard S, Balloux F. Discriminant analysis of principal components: a new method for the analysis of genetically structured populations. BMC Genet. 2010; 11: 94. pmid:20950446
- 40. Beerli P, Palczewski M. Unified framework to evaluate panmixia and migration direction among multiple sampling locations. Genetics. 2010; 185: 313–326. pmid:20176979
- 41. Palero F, Abello P, Macpherson E, Beaumont M, Pascual M. Effect of oceanographic barriers and overfishing on the population genetic structure of the European spiny lobster (Palinurus elephas). Biol J Linn Soc. 2011; 104: 407–418.
- 42. Muths D, Tessier E, Bourjea J. Genetic structure of the reef grouper Epinephelus merra in the West Indian Ocean appears congruent with biogeographic and oceanographic boundaries. Mar Ecol. 2015; 36: 447–461.
- 43. Liu J, Gao T, Yokogawa K, Zhang Y. Differential population structuring and demographic history of two closely related fish species, Japanese sea bass (Lateolabrax japonicus) and spotted sea bass (Lateolabrax maculatus) in Northwestern Pacific. Mol Phylogenet Evol. 2006; 39: 799–811. pmid:16503171
- 44. Liu J, Gao T, Wu S, Zhang Y. Pleistocene isolation in the Northwestern Pacific marginal seas and limited dispersal in a marine fish, Chelon haematocheilus (Temminck & Schlegel, 1845). Mol Ecol. 2007; 16: 275–288. pmid:17217344
- 45. Carlsson J. Effects of microsatellite null alleles on assignment testing. J Hered. 2008; 99: 616–623. pmid:18535000
- 46. Rousset F, Raymond M. Testing heterozygote excess and deficiency. Genetics. 1995; 140: 1413–1419. pmid:7498780
- 47. Was A, Gosling E, Hoarau G. Microsatellite analysis of plaice (Pleuronectes platessa L.) in the NE Atlantic: weak genetic structuring in a milieu of high gene flow. Mar Biol. 2010; 157: 447–462.
- 48. Waples RS. Testing for Hardy-Weinberg proportions: have we lost the plot? J Hered. 2015; 106: 1–19. pmid:25425676
- 49. Slate J, Pemberton J. Admixture and patterns of linkage disequilibrium in a free-living vertebrate population. J Evol Biol. 2007; 20: 1415–1427. pmid:17584236
- 50. Adams SM, Lindmeier JB, Duvernell DD. Microsatellite analysis of the phylogeography, Pleistocene history and secondary contact hypotheses for the killifish, Fundulus heteroclitus. Mol Ecol. 2006; 15: 1109–1123. pmid:16599970
- 51. Mallet J. A species definition for the modern synthesis. Trends Ecol Evol. 1995; 10: 294–299. pmid:21237047
- 52. Martínez P, Bouza C, Castro J, Hermida M, Pardo B, Sánchez L. Analysis of a secondary contact between divergent lineages of brown trout Salmo trutta L. from Duero basin using microsatellites and mtDNA RFLPs. J Fish Biol. 2007; 71: 195–213.
- 53. Bossu CM, Near TJ. Characterization of a contemporaneous hybrid zone between two darter species (Etheostoma bison and E. caeruleum) in the Buffalo River system. Genetica. 2013; 141: 75–88. pmid:23443763
- 54. Toews DP, Brelsford A. The biogeography of mitochondrial and nuclear discordance in animals. Mol Ecol. 2012; 21: 3907–3930. pmid:22738314
- 55. Wielgoss S, Gilabert A, Meyer A, Wirth T. Introgressive hybridization and latitudinal admixture clines in North Atlantic eels. BMC Evol Biol. 2014; 14: 61. pmid:24674242
- 56. Miralles L, Machado-Schiaffino G, Garcia-Vazquez E. Genetic markers reveal a gradient of hybridization between cape hakes (Merluccius capensis and Merluccius paradoxus) in their sympatric geographic distribution. J Sea Res. 2014; 86: 69–75.
- 57. Machado-Schiaffino G, Juanes F, Garcia-Vazquez E. Introgressive hybridization in North American hakes after secondary contact. Mol Phylogenet Evol. 2010; 55: 552–558. pmid:20138222
- 58. Montanari SR, Hobbs JA, Pratchett MS, Bay LK, Van Herwerden L. Does genetic distance between parental species influence outcomes of hybridization among coral reef butterflyfishes? Mol Ecol. 2014; 23: 2757–2770. pmid:24750170
- 59. Schunter C, Carreras‐Carbonell J, Macpherson E, Tintoré J, Vidal-Vijande E, Pascual A, et al. Matching genetics with oceanography: directional gene flow in a Mediterranean fish species. Mol Ecol. 2011; 20: 5167–5181. pmid:22097887
- 60. Astraldi M, Bianchi C, Gasparini G, Morri C. Climatic fluctuations, current variability and marine species distribution-a case-study in the Ligurian sea (north-west Mediterranean). Oceanol Acta. 1995; 18: 139–149.
- 61. Hobbs JP, Frisch AJ, Allen GR, Van Herwerden L. Marine hybrid hotspot at Indo-Pacific biogeographic border. Biol Lett. 2009; 5: 258–261. pmid:19126528
- 62. Damerau M, Matschiner M, Salzburger W, Hanel R. Comparative population genetics of seven notothenioid fish species reveals high levels of gene flow along ocean currents in the southern Scotia Arc, Antarctica. Polar Biol. 2012; 35: 1073–1086.
- 63. Durand J, Collet A, Chow S, Guinand B, Borsa P. Nuclear and mitochondrial DNA markers indicate unidirectional gene flow of Indo-Pacific to Atlantic bigeye tuna (Thunnus obesus) populations, and their admixture off southern Africa. Mar Biol. 2005; 147: 313–322.
- 64. Teacher AG, André C, Jonsson PR, Merilä J. Oceanographic connectivity and environmental correlates of genetic structuring in Atlantic herring in the Baltic Sea. Evol Appl. 2013; 6: 549–567. pmid:23745145
- 65. Kai Y, Sakai K, Orr JW, Nakabo T. Secondary contact in the Sea of Japan: the case of the Careproctus rastrinus species complex (Liparidae). Ichthyol Res. 2011; 58: 366–369.
- 66. Mallet J. Hybridization, ecological races and the nature of species: empirical evidence for the ease of speciation. Philos Trans R Soc Lond B Biol Sci. 2008; 363: 2971–2986. pmid:18579473
- 67. Yamada H. Hatching and early feeding of Japanese sand eel Ammodytes personatus in the Ise bay [Japan]. Nippon Suisan Gakkaishi. 1998; 64: 440–446.