High genetic diversity and low differentiation in mud crab (Scylla paramamosain) along the southeastern coast of China revealed by microsatellite markers

The mud crab (Scylla paramamosain Estampador 1949), mainly distributed along the southeastern coast of China, is a commercially important crab resource for fisheries and aquaculture. Records of S. paramamosain aquaculture date back more than 100 years in China (Shen and Lai, 1994) and more than 30 years in other Asian countries (Keenan and Blackshaw, 1999). In wild environments, adults mate inshore and the gravid females generally migrate offshore to spawn (Perrine, 1979). Because of over-exploitation and environmental deterioration, numbers in the wild have decreased quickly. In order to conserve and sustainably harvest this important crab resource, genetic studies are necessary as they enable a better understanding of genetic diversity and structure (Dickerson et al., 2010), allow investigation of phylogenetic and evolutionary history (Gvozdík et al., 2010; Van Syoc et al., 2010), and also provide constructive guidance for resource conservation and management (Ortega-Villaizan Romo et al., 2006). A mitochondrial DNA (mtDNA) study of S. paramamosain indicated a genetically homogeneous population structure and a recent population expansion event (He et al., 2010). Moreover, again using mtDNA, a high level of genetic diversity and low genetic differentiation at different locations were observed in S. paramamosain inhabiting the southeastern coast of China (Lu et al., 2009; Ma et al., 2011a).

Microsatellites are nuclear molecular markers characterized by a 1–6 bp length repeat motif, high polymorphism and co-dominant inheritance. Microsatellite markers have been widely used for investigation of genetic diversity (Dudaniec et al., 2010), determination of pedigree (Li et al., 2009a), construction of genetic maps (Ma et al., 2011b) and mapping of quantitative trait loci (Zhang et al., 2011). To date, microsatellite markers have been isolated in S. paramamosain (Takano et al., 2005; Ma et al., 2010; Ma et al., 2011c; Cui et al., 2011), but no information about population genetic diversity and differentiation has been reported for this important crab species.

In this study, a total of 397 wild specimens from 11 locations on the southeastern coast of China were sampled and genotyped using nine microsatellite markers. The purpose was to investigate the level of population genetic diversity and differentiation in S. paramamosain across these regions to provide valuable information for conservation, harvesting and management of this key fishery resource.

A total of 397 wild specimens of S. paramamosain were collected from the following locations along the southeastern coast of China: Sanmen (SM, N=38), Ningde (ND, N=35), Zhangzhou (ZZ, N=32), Shantou (ST, N=25), Shenzhen (SZ, N=40), Zhanjiang (ZJ, N=41), Haikou (HK, N=37), Wenchang (WC, N=51), Wanning (WN, N=35), Dongfang (DF, N=30) and Danzhou (DZ, N=33) (Fig. 1; Table 1). Each specimen was killed by a lethal dose of MS-222. Genomic DNA was extracted from muscle tissue using traditional proteinase K and phenol–chloroform extraction protocols as described previously (Ma et al., 2009). The DNA was adjusted to 100 ng μl^–1 and stored at –20°C until use.

Nine polymorphic microsatellite loci were selected for genotyping, of which eight were developed using the 5′ anchored PCR method (Cui et al., 2011), and the remaining one was developed using PCR-based isolation of microsatellite arrays (PIMA) (Ma et al., 2010) in our laboratory (Table 2). The criteria for selection were as follows: annealing temperature of 50–63°C, expected product size between 110 and 320 bp, observed heterozygosity value >0.5 and no stuttering bands. PCR reactions were conducted in a total volume of 25 μl and included 0.4 μmol l^–1 each primer, 0.2 mmol l^–1 each dNTP, 1× PCR reaction buffer, 1.5 mmol l^–1 MgCl₂, 0.75 U Taq polymerase and approximately 100 ng template DNA, under the following conditions: one cycle of denaturation at 94°C for 4 min; 30 cycles of 30 s at 94°C, 50 s at a primer-specific annealing temperature (Table 2), and 50 s at 72°C. As a final step, products were extended for 7 min at 72°C.

Several methods, including agarose gel electrophoresis, denaturing polyacrylamide gel electrophoresis and automated DNA sequencing, were employed for detecting differences in nucleotide sequence, of which the second is a very effective and practical technique for genotyping of microsatellites and has been used in a wide range of organisms, as it has many advantages: a high resolution (about 1 bp) and large output (100 samples each), low expense and easily mastered. In this study, the PCR products were separated on 6% denaturing polyacrylamide gels as described previously (Ma, 2009). The microsatellite fragments were visualized by silver staining, which was performed as follows. The gel was soaked in 1.0 l staining solution (1.5 g AgNO₃) for about 10 min; this solution was then removed and the gel was washed in ddH₂O for 5 s. Then the gel was soaked in 1.0 l coloured solution (20 g NaOH and 4 ml formaldehyde) for about 10 min. Finally, the gel was cleaned with ddH₂O. The size of alleles was estimated according to the pBR322/Msp I marker.

Observed and expected heterozygosity, departure from Hardy–Weinberg equilibrium (HWE), linkage disequilibrium (LD) and inbreeding coefficient (F_IS) were obtained using ARLEQUIN version 3.01 software (Excoffier et al., 2005). Genetic differentiation among locations was estimated using the analysis of molecular variance (AMOVA) approach by GENAlEX version 6.41 software (Peakall and Smouse, 2006). The significance levels were tested by 10,000 permutations for LD and by 1000 permutations for fixation index (F_ST) values. Observed number of alleles (N_a), effective number of alleles (N_e) and genetic distance were estimated using POPGENE version 1.31 software (Yeh et al., 1999). An unweighted pair-group mean analysis (UPGMA) tree was constructed based on Nei's genetic distance (Nei, 1978) of pairwise locations using MEGA version 4.0 software (Tamura et al., 2007). The association between genetic differentiation and geographic distance (isolation by distance) among locations was estimated by the Mantel test (Mantel, 1967) with 1000 permutations.

All nine microsatellite loci used in this study were polymorphic in each location, showing a high level of genetic diversity (Table 3). In total, 104 alleles were detected from 397 individuals in 11 locations across nine loci. N_a ranged from six (Scypa1) to 16 (Scypa8 and Scpa03) per locus and from 7.8 (ST) to 9.6 (WC) per location. H_O and H_E ranged from 0.32 to 1.00 and from 0.31 to 0.93 per locus–location combination, and from 0.62 (SM) to 0.77 (HK) and from 0.66 (ST) to 0.76 (ND and DZ) per location, respectively. F_IS ranged from –0.278 to 0.440 per locus–location combination and from –0.137 (ST) to 0.136 (SM) per location, with an average of 0.001 as a whole.

An exact probability test of HWE was performed among 99 locus–location combinations, and it revealed a significant deviation at 19 loci (P<0.05). These 19 loci were Scypa1 (in ZZ and HK), Scypa2 (in SM and ZJ), Scypa3 (in ND and ST), Scypa4 (in DF), Scypa8 (in ND, ZZ, HK and WN), Scypa13 (in SM, ST, ZJ and WN) and Scpa03 (in SM, SZ, ZJ and WC). Two loci (Scypa5 and Scypa11) were in keeping with HWE in all locations. Probability tests of genotypic LD for all pairs of loci within each location suggested significant non-random associations in only one of 396 possible pairwise comparisons after sequential Bonferroni correction (Scypa2 and Scypa13 in DF, P<0.00139) (Rice, 1989). When each location was analysed separately, there was no evidence of stuttering and large allelic dropout in any of the loci, as confirmed by MICRO-CHECKER version 2.2.3 software (Van Oosterhout et al., 2004).

The AMOVA showed that genetic variation existed mainly within locations, rather than among locations, as the percentage of variance was 98.17% within locations and 1.83% among locations. Although the overall F_ST value for all locations and loci was statistically significant (F_ST=0.0183, P<0.05), the genetic differentiation was still low, because the F_ST value was much lower than 0.05 (Tables 4, 5). Multi-locus estimates of F_ST for all possible pairwise locations ranged from 0.002 (ZJ and DZ) to 0.067 (SM and ST). The highest differentiation was between SM and ST (F_ST=0.067), and the lowest differentiation was between ZJ and DZ (Table 5). Thirty-nine out of 55 pairwise locations showed significant differentiation (P<0.05). Nei's genetic distances between pairwise locations ranged from 0.0121 (ZZ and SZ) to 0.2036 (SM and ST), and were lower than 0.1 in 43 out of the 55 pairwise locations. Of the 11 locations, SM was the most distinctive, as it showed significant differentiation in relation to all the other 10 locations (F_ST values ranged from 0.024 to 0.067). In contrast, DZ was the most representative as it significantly differed from only four locations (F_ST values ranged from 0.012 to 0.029).

Cluster analysis of 11 locations using the UPGMA approach revealed two groups: one contained 10 locations and the other contained only one location (SM) (Fig. 2). Mantel tests for isolation by distance among locations detected a significant positive correlation between pairwise F_ST/(1–F_ST) and the natural logarithm of geographic distance (km) (r²=0.1139, P=0.02), while there was no significant correlation between pairwise F_ST and geographic distance (km) (r²=0.1230, P=0.06) (Fig. 3)

The results of this study suggest a high level of population genetic diversity of S. paramamosain along the southeastern coast of China (N_a, H_O and H_E in the range 7.8–9.6, 0.62–0.77 and 0.66–0.76 per location, respectively), in accordance with previous studies that showed a high level of mtDNA genetic diversity in S. paramamosain (Lu et al., 2009; Ma et al., 2011a). High population genetic diversity has also been observed in other marine animals, such as scallop (Chlamys farreri) (Zhao et al., 2009), Atlantic salmon (Salmo salar) (Karlsson et al., 2010) and silver pomfret (Pampus argenteus) (Zhao et al., 2011). Three factors including the life history characteristics, environmental heterogeneity and large population size may help to maintain a high level of genetic diversity (Perrine, 1979; Nei, 1987; Avise, 1998). On the whole, the level of genetic diversity of S. paramamosain from the southern regions was higher than that from the northern regions (Table 3), which may be due to the different environments. A similar finding was observed in a previous study, which indicated a trend for a reduction in genetic diversity of S. paramamosain from south to north, step by step using mtDNA (Lu et al., 2009).

Generally, marine fishes are considered to have a low level of genetic differentiation among different geographic populations because of the high dispersal capabilities, large population sizes and relatively small barriers in the marine environment (Beheregaray and Sunnucks, 2001). For the fish Nibea albiflora, there was little difference in the population genetic structure between the Yellow Sea and East China Sea observed using mtDNA (Han et al., 2008). For the shrimp Fenneropenaeus chinensis, no significant population genetic differentiation between the Yellow Sea and Bohai Sea was found using both microsatellite DNA and mtDNA (Liu et al., 2006; Li et al., 2009b). For the crab S. paramamosain, a genetically homogeneous population structure with high gene flow was observed among most localities along the coasts of the East China Sea and South China Sea using mtDNA (He et al., 2010). In the current study, statistically significant genetic differentiation was detected across 11 locations along the southeastern coast of China (F_ST=0.0183, P<0.05), but the F_ST value was still low (<0.05), suggesting a low level of genetic differentiation (Wright, 1978). A similar finding was reported in earlier studies, which suggested a low differentiation in S. paramamosain using mtDNA (Lu et al., 2009; He et al., 2010; Ma et al., 2011a). The above information indicates that all locations of S. paramamosain should be a single genetically homogeneous population. The low F_ST value indicates a relatively high gene flow among locations. There are three probable explanations for this: (1) the unique reproductive habit in which adults and juveniles migrate between ocean basins and adjacent continental margins; (2) the high dispersal capabilities of larvae; and (3) the relatively small physical barriers in the marine environment.

Among these 11 locations, SM was the most genetically distinctive in two main ways: (1) it has the lowest genetic diversity (the overall H_O was 0.62) and the highest F_ST values (between 0.024 and 0.067) compared with other locations; and (2) it has the greatest overall F_IS value (F_IS=0.136, P<0.05) compared with other locations. These findings indicate that the gene exchange is relatively low between SM and other locations compared with that between other location pairs. The optimum temperature range of this crab is 18–27°C for growth, and a higher temperature is needed for spawning. However, SM is the most northern of these locations, so the seawater temperature is the lowest in the same period. Low temperature may limit the effective population size and the high dispersal capabilities of S. paramamosain. Over-fishing by humans may be another potential explanation. A significant positive correlation between genetic differentiation and geographic distance was found, suggesting an isolation by distance model of genetic variation.

In conclusion, a high level of population genetic diversity and low differentiation were found in the mud crab (S. paramamosain) from 11 locations along the southeastern coastal regions of China by microsatellite analysis, which showed a genetically homogeneous population structure for S. paramamosain in these 11 locations. In the future, more population genetic studies should be carried out in this crab species. The findings in this study will provide valuable information for conservation, harvesting and artificial selective breeding of this important fishery resource.

 
• AMOVA

  analysis of molecular variance

 
• F_IS

  inbreeding coefficient

 
• F_ST

  fixation index

 
• H_E

  expected heterozygosity

 
• H_O

  observed heterozygosity

 
• HWE

  Hardy–Weinberg equilibrium

 
• LD

  linkage disequilibrium

 
• N_a

  observed number of alleles

 
• N_e

  expected number of alleles

We thank Dr Keji Jiang for assistance in partial sample collection.