Genetic Analysis of Cultured and Wild Populations of Mytilus coruscus Based on Mitochondrial DNA

: DNA sequences from the mitochondrial gene cytochrome oxidase subunit I (mtDNA CO I ) were used to estimate the genetic variability in two wild populations and two cultured populations of the hard shelled mussel, Mytilus coruscus . Thirty haplotypes were identified in the four populations. The cultured populations exhibited a lower number of haplotypes and genetic diversity than those of the wild populations, suggesting that a small number of effective founding breeders contributed to the genetic variation of the cultured populations. No significant differentiation was observed between the cultured population and local wild population, implying that persistent gene flow occurred in these populations. This genetic survey is intended as a baseline for future genetic monitoring of M. coruscus aquaculture stocks.

The hard shelled mussel, Mytilus coruscus, is an economically important mussel which is widely distributed from coast of China, Japan, and Korea (Wang, 1997). It is documented that this species occurred from Dalian to Xiamen in China. The culture of M.coruscus has been carried out in only a few regions of China in the past decades. The cultured juveniles mainly originated from the collection of natural populations. In recent years, natural juveniles have decreased while mussel farmers have increased. Due to overexploitation, most mussel stocks have dramatically declined. Therefore, it is very important to obtain juveniles that are produced by artificial breeding to increase the juvenile population.
The first hatchery stock is being developed in Zhejiang province, China. However, differences in allele frequencies between farmed strains and the wild source populations have been shown to be a result of breeding related individuals or the use of small numbers of individuals as brood-stock, leading to lower genetic variability of the farmed strains (McGinnity et al, 2003). Additionally, farmed strains of small population size are more sensitive to genetic drift and consequently have lower within population genetic diversity than wild populations (Allendorf, 1986). Genetic variability is the primary resource in the successful artificial propagation of any species. Proper management and breeding programs must be implemented to preserve genetic variability, and prevent inbreeding depression. However, for such programs to be successful, information on the genetic relationships among cultured and wild populations is essential.
Mitochondrial DNA has been used extensively for studies in molecular ecology (Avise, 2000). In comparison to allozyme or nuclear DNA, the higher mutation rate, smaller effective population size is expected to provide greater power to identify population structure. Additionally, mtDNA gives a better estimate of genetic differentiation than nuclear markers since it is approximately fourfold more sensitive to genetic drift and founder effects (Birky et al, 1983). Mitochondrial Cytochrome Oxidase I gene (mtCOI) has been used widely in marine species, such as Tegillarca granosa (Zheng et al, 2009)，due to adequate levels of variability and easy amplification via universal primers (Folmer et al, 1994).
In this study, we used the mitochondrial Cytochrome Oxidase I gene (mtCOI) to describe the genetic variability of two cultured, and two wild populations and also to quantify the genetic differentiation between them.

Samples
Specimens of M.coruscus obtained from both cultured and natural populations are listed in Tab. 1 and Fig. 1. The two wild populations were randomly sampled by divers from the subtidal zone of Shengsi and Lianjiang in 2007，respectively (coded WSS and WLJ). The two cultured populations were of Shengshan (coded CSS) and Huaniao (coded CHN), and also collected in 2007. The CSS population was collected from the cultured area of M.coruscus. In this area, the aquaculture practice had been carried out for three decades. A large number of the natural spat collected were commonly reared. The CHN population hatchery stock was the first-generation of offspring produced in spring 2005 using hundreds of wild caught M.coruscus in Shengsi (exact numbers not documented). Gill tissues were obtained and stored in 100% ethanol at room temperature until DNA extraction.

Extraction of DNA
Genomic DNA was isolated from gill using the standard proteinase K digestion and phenol/chloroform extraction procedures described by Shen et al (2006). DNA quality was assessed by running samples on 1% agarose gels, and DNA concentration was measured with an UV/visible spectrophotometer (Eppendorf AG 22331 Hamburg) for absorption at 260 nm. DNA was diluted to 50 ng/μL for polymerase chain reaction (PCR) amplifications.

Amplification of DNA and sequencing
PCR was carried out in 50μL reactions containing 100 ng DNA sample, 5 μL 10×PCR buffer, 2.0 μmol/L MgCl 2 , 200 μmol/L dNTPs, 0.2 μmol/L primer, 1U Taq polymerase. A pair of universal primers was used in this study (forward primer LCO1490：5'-GGTCAACAA-ATCATAAAGATTGG-3'; reversal primer HCO2198： 5'-TAAACTTCAGGGTGACCAAAAAATCA-3' (Folmer et al, 1994). PCR was performed using the Eppendorf PCR system programmed for an initial denaturation step of 3 min at 94℃ followed by 35 cycles, each consisting of 94℃ denaturation for 50 second, 48℃ annealing for 50 second and 72℃ extension for 1 min. A final extension of 10 min was performed at 72℃ and the PCR products were then held indefinitely at 4℃. A negative control, consisting of all the reaction components except template DNA, was also included for each of amplification.
PCR products were visualized using 1.5% agarose gel stained with ethidium bromide. All amplified products were purified using TIANquick Midi Purification Kit (Tiangen, China). Purified PCR products were directly sequenced in both directions using the PCR primers on an Applied Biosystems ABI 3730 DNA sequencer.

Data analysis
For all sequence analyses, sequences were aligned with BioEdit Sequence Alignment Editor version 7.0.9

Zoological Research
Vol. 30 Fig. 1 Sample locations of two wild and two cultured Mytilus corcuscus (Hall, 1999) and saved in the Fasta format. The identical haplotypes in the aligned matrix were identified and collapsed using COLLAPSE version 1.2 (Posada, 2006). Nucleotide composition was computed in MEGA version 4.0 (Tamura et al, 2007). Genetic diversity among wild and cultivated populations was estimated using haplotype number (N), gene diversity (h), and nucleotide diversity (π) as implemented in DnaSP version 4.0 (Rozas et al, 2003). Gene diversity was calculated according to Nei (1987), using the probability that two randomly chosen haplotypes of the sample are different. Nucleotide diversity was also computed according to Nei (1987), as the probability that two randomly chosen homologous sites are different. Genetic diversity was compared between farmed and wild populations of mussel using FSTAT (Goudet, 1995). Pairwise estimates of F st and corrected average pairwise population distances among all populations were obtained using the program Arlequin3.1 (Excoffier et al, 2005). The significance of these estimates was tested by comparing observed F st with a null distribution obtained by 10 000 random permutations of the data set (Excoffier et al, 1992).

Characteristics of mtDNA COI of M.coruscus
Once all sequences were aligned, a total segment length of 573 bp of cytochrome oxidase subunit I (COI) was obtained. The average nucleotide composition in our samples is A=33.6%, C=23.9%, G=15.8%, T=26.6%, with an average G+C content of 39.7%. These sequences include 59 variable sites, of which 41 are singleton variable sites and 18 are parsimony informative sites. Sequences have been deposited in GenBank under the following accession number range: FJ495257-FJ495286.

Genetic variability and population structure within populations
The samples examined show a wide range of values for the number of haplotypes, gene diversities and nucleotide diversities (Tab. 2). A total of 30 haplotypes were obtained for wild and cultured populations of M. coruscus (Tab. 3). The number of individuals of different haplotypes ranged from 1 to 18. The haplotype Hap2 had the greatest number of individuals across all samples. The wild WSS population had the greatest number of haplotypes, whereas the cultured CHN population originating from the Shengsi area had the fewest number of haplotypes. Every population had population-specific haplotypes. In the cultured WLJ population, only 3 specific haplotypes were found, exhibiting a decrease of specific haplotypes. In contrast, the samples from the wild WSS population exhibited the greatest number of population-specific haplotypes. The samples from natural populations exhibited the greatest number of haplotypes and population-specific haplotypes compared to cultured samples. The samples from wild WSS population had an H value of 0.911 and π of 0.00898, compared to the aquaculture samples CHN and CSS with H values of 0.718, 0.911 and π values 0.00293, 0.00834, respectively. The wild WSS population had 3-fold higher nucleotide diversity than the cultured CHN population obtained through artificial propagation. All cultured populations exhibited the less genetic diversity for gene diversity and nucleotide diversity compared to the wild populations.

Genetic differentiation among populations
Tab .

Discussion
The artificial breeding program for this species starts for a short time and will expand accordingly. There is no previous genetic record of this species and no attempt has been made to assess the genetic status of both wild and cultured populations of this species. This is the first study to demonstrate that a mitochondrial marker can be used to monitor changes in the genetic diversity of the hard-shelled mussel (M.corcuscus) during domestication.
Genetic variability is an important attribute of species under domestication, since those with higher levels of variation are most likely to present high additive genetic variance for production traits (Alarcon et al, 2004). In this study, genetic analysis of mitochondrial COI gene sequences revealed higher genetic variability for gene diversity and nucleotide diversity both in wild and cultured M. corcuscus than for the same sequence regions in other marine organism. Results from mitochondrial COI gene sequences in crayfish showed nucleotide diversity ranging from 0.01% to 0.43% (Trontelj et al, 2005). In the mud crab Scylla serrata, the COI sequence divergence ranges from 0.17 % to 0.46% and gene diversity ranges from 0.37 to 0.85 (Fratini & Vannini, 2002). Relatively high levels of DNA diversity characterized the population of the bivalve mollusc Congeria kusceri (haplotype diversity=0.66 in the COI gene) (Stepien et al, 2001). However, the mtDNA nucleotide and gene diversity were reduced in the cultured population. A slight decrease of genetic variability in cultured populations had been observed in many fish and mussel species (Lundrigan et al, 2005;Pampoulie et al, 2006;Wang, 2007;Kong & Li, 2007;Shu et al, 2008). The reduced genetic variability we observed in the farmed strains is probably due to a low number of successful breeders during the foundation period, similar to a recent bottleneck in terms of impact on genetic variability (Allendorf, 1986). The greatest reductions in number of haplotypes, nucleotide diversity and gene nucleotide observed in cultured CHN population in this study are probably caused by the use of small numbers of brood-stock collected from wild populations. The reductions in the cultured CSS population are probably contributed to collection of natural spat from small numbers of parents and samples examined. A more precise assessment of the genetic variability in cultured populations can be made, were the magnitude of the genetic variation of wild populations made available. This is almost a requirement when the number of generations of the cultured populations is small and there has not been enough time for the variation to be reduced to a detectable level.

Tab. 3 mtDNA haplotypes distribution of Mytilus coruscus populations (GenBank accession no and number of individuals)
The comparison of the F st values (Tab. 4) between CSS, CHN and WSS populations showed that the values were not significantly different. This indicates that no significant genetic differentiation was detected between CHN, CSS and WSS populations. Similar findings were reported for fish (Yang et al, 2008), marine mussel (Jiang et al, 2007;He et al, 2008). The slow genetic differentiation might be due to persistent gene flow caused by cultured juvenile practice in open sea, close to the wild population. But significant genetic differentiation observed between WLJ and WSS, CHN populations, was probably associated with geographic factors. This regional variation could be an important source of diversity for both genotypic and phenotypic traits to be selected in accordance with aquaculture goals (Lundrigan et al, 2005).
Large scale approaches to assess the genetic structure of wild and cultured populations, such as those reported here for M. coruscus, are most suitable to provide insights about the evolutionary history of a population and its potential as a source of variation for cultured stock. More detailed information obtained in future studies will be used as a baseline for genetic monitoring of M.coruscus aquaculture stains, in particular, the first-generation, recently cultured population. Through population management, inbreeding can be prevented through the maintenance of population size above a critical level, with regular observation of heterozygosity levels and breeding to maximize genetic variability and minimize inbreeding (Barroso et al, 2005). Future farming practice should avoid culturing in areas in which local wild population inhabit, since wild populations represent the primary source of genetic variability for aquaculture stocks. This genetic survey is intended as a baseline for future genetic monitoring of M.coruscus aquaculture stocks.