Genetic diversity and population structure of Meretrix petechialis in China revealed by sequence-related amplified polymorphism markers

Collection of samples and extraction of DNA

In this study, samples were collected from different tidal flats. A total of 270 samples were obtained from nine locations. We collected the clam M. petechialis at Fujian (FJ), Guangxi (GX), Panjin (PJ), Shandong (SD), Dandong (DD), Hainan (HN), Dalian (DL), North Korea (NK) and Jiangsu (JS) (Fig. 1). The adductor muscles were dissected from the samples and stored in 100% ethanol. DNA was extracted from the adductor muscles using a TIANamp Marine Animals DNA Extraction Kit (Tiangen, Beijing, China). After successful extraction, concentration was determined with 1% agarose gel electrophoresis and the quality of the extracted DNA was determined. The extracted DNA was stored in TE buffer and PCR was performed after the quality was confirmed. The diluted concentration was 100 ng/μL.

Optimization of SRAP–PCR amplification system and detection of products

Primers for Me7 and Em15 were based on published SRAP sequence (Table 1), and 105 pairs of primers were obtained by crossing the upstream and downstream. The parameters of the PCR were as follows: 2.5 μL of 10× PCR buffer; two μL dNTP mixture; 0.25 μL Taq DNA polymerase produced by Takara, Japan; one μL upstream and downstream primers (10 μmol/L), and one μL template DNA, ddH₂O to make up the volume of the PCR reaction to 25 μL. The reaction conditions were: 5 min at 94 °C, 5 cycles, 1 min at 94 °C, an annealing temperature of 35 °C for 1 min, an extension temperature of 72 °C, and an extension time of 1 min, 1 min at 94 °C, an annealing temperature of 50 °C for 1 min, extension at 72 °C for 1 min, a total of 35 cycles; extension at 72 °C for 10 min and stored at 4 °C. After the PCR was completed, it was detected by 6% denatured polyacrylamide gel electrophoresis. To ensure the accuracy of denatured polyacrylamide gel electrophoresis, a control was used for each test.

SRAP band statistical analysis and calculation of genetic indicators

According to the SRAP amplification electrophoresis procedure, each of the bands is considered to be a locus, with reference to 10 bp DNA markers. The SRAP of each sample amplification band is counted, with a clearly visible band recorded as 1 and no band recorded as 0, and converted into 1 and 0 matrices. The polymorphism information content (PIC) value of the primer is calculated using PIC software. POPGENE32 (Yeh, Yang & Boylet, 1999) was used to observe the number of alleles (NA), the effective number of alleles (Ne), the Shannon’s information index (I), the Nei’s gene diversity (H), the number of polymorphic loci (N), and the percentage of polymorphic loci (P), the Nei’s unbiased genetic distance (D) and genetic similarity (S) were calculated and the neighbor-joining (NJ) tree of nine M. petechialis groups was constructed according to the genetic distance with MEGA 7.0 (Kumar, Stecher & Tamura, 2016). The total genetic variation among the samples was calculated using the phi-statistic through the analysis of molecular variance (AMOVA). This analysis was performed using the computer program GENALEX ver.6.5 (Peakall & Smouse, 2006). The total genetic variation is partitioned at three levels—within populations (Phi-PT), among populations within regions (Phi-PR) and among regional populations (Phi-RT) (Xu et al., 2008).

The patterns of the population structure were further investigated using the model-based Bayesian clustering procedure in STRUCTURE version 2.3.4 (Pritchard, Stephens & Donnelly, 2000), the data was converted into the appropriate format and imported into the software and relevant parameters were set. We assigned individuals to K populations based on their multilocus genotype. The estimated K value was set to 1–9 then STRUCTURE was used with 10,000 iterations and a burn-in period of 10,000. All runs were repeated 10 times at each K, the most appropriate value of K was identified with calculated ΔK by using Structure Harvester ver.0.6.92 (http://taylor0.biology.ucla.edu/structureHarvester/) (Earl & Vonholdt, 2012). Then, the optimal K value was independent run by the CLUMPP 1.1.2 (Jakobsson & Rosenberg, 2007), and the output was used as input into DISTRUCT 1.1 (Rosenberg, 2004) for cluster visualization. Furthermore, Principal Component Analysis (PCA) was employed to assess the degree of genetic relatedness among groups using Origin Pro.

SRAP amplification band type

Using the genomic DNA of nine populations of M. petechialis, 105 pairs of SRAP primer combinations were screened, and 30 pairs of primer combinations with clear, stable and few bands were selected, of which eight pairs of primer combinations (Me2/Em15, Me3/Em4, Me4/Em3, Me4/Em5, Me4/Em11, Me5/Em5, Me5/Em14, Me6/Em5) can obtain stable, reproducible and polymorphic amplification maps. The polymorphic loci amplified by different primer combinations are different. The number of polymorphic loci detected by each pair of primers is 12–23, and the average number of loci per pair of primers is 16.5. The fragment size range of 116–250 bp has a high polymorphism, which indicates that SRAP can be applied to study genetic diversity in M. petechialis.

Population genetic diversity

A total of 132 polymorphic loci were detected in eight pairs of primer combinations. The total stripe number was 308, of which the Me5/Em14 polymorphic stripe number was the highest (23). The Me3/Em4 primer’s polymorphic frequency was the highest for 53.85% and the Me6/Em5 was the lowest at 29.27% (mean = 42.86%). The PIC value of the Me5/Em14 primer was the highest (0.90), the PIC value of the Me4/Em5 primer was the lowest (0.78) and the average PIC value of eight pairs of primers was 0.86 (Table 2). The highest number of polymorphism bits and percentages of polymorphism sites in the JS population were 36% and 27.27% and the lowest in DL and NK populations were 32% and 24.24%, respectively. The Nei’s gene diversity of the nine M. petechialis populations ranged from 0.0647 (SD) to 0.0793 (NK) and the Shannon’s information index was between 0.1023 and 0.1202, with the SD population being the lowest and the JS population being the highest (Table 3).

Population genetic differentiation

The Nei’s unbiased measures genetic identity and genetic distance of nine geographical groups of M. petechialis were calculated according to the allele frequency of each locus and the results were shown in Table 4. The genetic distance between the GX and SD groups was the largest (0.0570) and the genetic similarity was the smallest (0.9446); The genetic distance between the SD and JS groups were the closest (0.0243) and the genetic similarity was the largest (0.9760). The cluster analysis of neighbor-joining trees showed that the GX forms a single population (Fig. 2), indicating that the genetic distance of the GX population was the furthest and there was a certain genetic differentiation.

Population clusters and structure analysis

According to the method of Earl & Vonholdt (2012), we used structure harvester website to analysis the K, the program provides a fast way to assess and visualize likelihood values across multiple values of K and thousands of iterations for easier detection of the number of genetic groups that best data. In the structure analysis, the highest ΔK = 8 (Fig. S1), which suggested that the nine populations of M. petechial were split into eight clusters (Fig. 3) and from the overall distribution of colors in it, we saw that most of the background colors of the DL and NK populations are the same, suggesting they were formed the same cluster, we suspected that it might be due to the close distance and similar genetic backgrounds. In contrast, the pink color which represented the GX group was hardly appeared in any other group, indicating that GX population has single genetic background, distant relationship and almost no gene flow with others populations. Additionally, the color composition of the remaining groups showed admixture with other populations color, it suggested that have a complex background and there is a certain degree of gene flow between the populations. The result was similar using the neighbor-joining dendrogram (Fig. 2). This also revealed that the northern and southern populations were not very integrated and their distribution pattern was could be based on geographical distance. To further elucidate the gene differentiation between populations, we performed (PCA) using Origin Pro. PCA was a supplement to cluster analysis, which analyzed the structural similarity of germplasm resources at marker locus. The PCA of 270 samples was calculated, in 132 principal components, the cumulative contribution rate of the first two principal components was 12.1%, contribution rates of the first and second principal components were 6.5% and 5.6%, respectively (Fig. 4). Nine populations were distributed into four different clusters, the PCA diagram revealed one cluster was GX population, two cluster included FJ, PJ, DD and three clusters included SD, HN, JS and four included DL, NK. In general, the classification results of each cluster have high correspondence with the NJ tree and structure. And there are many overlapping phenomena among samples in different geographical areas, but no obvious regional difference. However, because the cumulative contribution rate of the first two principal components is very low, which can reflect less genetic information and cannot accurately explain the genetic diversity information among populations. In this study, the AMOVA results were obtained from nine populations of M. petechial (Table 5). It revealed that populations within regions contributed to 10% of the total genetic variance (Phi-PR). The genetic variance within populations (Phi-PT) was 37% and 53% of the total genetic variance (PhiRT) was calculated between populations from different regions. This indicated that there was significant genetic differentiation within populations. The results obtained from the Nei’s genetic diversity analysis were consistent. The gene flow estimate obtained by PhiPT among populations (Nm (Haploid) = [(1/PhiPT)−1]/2) was 0.555, the Nm value was less than 1.0 and the number of gene exchanges between each generation was 0.555.