1 INTRODUCTION

The edible jellyfish, Rhopilema esculentum Kishinouye, a cnidarian of the class Scyphozoa, order Rhizostomeae, and family Rhopilema is an important fishery resource for its high economic value (Hong, 2002). It has been a valued seafood in china for thousands of years and jellyfish-related businesses generate millions of dollars each year in Asia. For example, it has been estimated that approximately 5 400-10 000 tons of jellyfish products are imported per year in Japan (Omori and Nakano, 2001).However, the jellyfish fishery is characterized by considerable fluctuations in catch and the natural resources of R. esculentum have significantly decreased in recent decades with the deterioration of the marine environment in China. R. esculentum has a metagenetic life cycle with asexual generation (strobilation) of polyps that then produce medusae via sexual reproduction (Ding and Chen, 1981). R. esculentum grow rapidly and ephyrae develop into sexually mature medusae in only three months. Therefore, the jellyfish is regarded as an ideal artificial seeding and releasing species. Since the 1980s, researchers have carried out jellyfish releasing experiments in Shandong and Liaoning Provinces in China, achieving remarkable success (Zhang, 2013).

Nevertheless, there are still several problems need to be considered. First, farming production and fishing yields are unstable. The jellyfish resource has declined annually since 2006 with the deterioration of the marine environment (You et al., 2012). These fluctuations are bound to influence the genetic diversity of populations, which can lead to genetic bottlenecks. Second, research on releasing has mainly focused on the catching yields of jellyfish (Zhang, 2013), while the genetic diversity of hatchery released populations has not been studied deeply. R. esculentum is dioecious and has a high reproductive capacity. In the process of farming, the number of parents is limited, and the males and females have very similar characteristics. Inbreeding depression can easily occur in a population if no attention is paid to the selection and collection of parents (Ma et al., 2005). This, in turn, will affect offspring growth and survival. Therefore, it is important to analyze the genetic diversity of hatchery populations before release. Third, the recapture rate is low and there is no accurate assessment method. According to statistics, the highest and lowest recapture rates in Liaoning are only 1.02% and 0.07%, respectively, with an average of 0.68%. In Zhejiang Province, the rate ranges from 0.57% to 2.30% (mean=1.34%), which indicates that R. esculentum recapture has enormous potential (Liang et al., 2007). Tagging jellyfish is difficult because their bodies are soft, transparent, and lightcolored. The accurate evaluation of the recapture rate requires urgent investigation; its absence restricts the hatchery release research process considerably.

Microsatellites are codominant markers that are common in the genome. They are highly polymorphic, stable, and adhere to Mendelian inheritance laws. These markers have been used in aquatic animal breeding and other studies, such as the construction of genetic linkage maps (Ruan et al., 2010), genetic diversity analysis (Hulak et al., 2010), and parentage identification (Borrell et al., 2011), among others. Microsatellite markers can be successfully used to identify either individuals or families based on traits that conform to Mendelian inheritance. To date, this technique has been validated in Chinese shrimp (Fenneropenaeus chinensis) (Wang et al., 2014a), Japanese flounder (Paralichthys olivaceus) (Sekino et al., 2003) and many other aquatic animal enhancement programs. However, there are currently no reports on R. esculentum in the literature. In the present study, a set of ten microsatellite markers were used to investigate the levels of genetic diversity and inbreeding within an R. esculentum hatchery population; variation in clones from the same scyphistomae was also assessed. The results of this paper reveal genetic variation in a hatchery population and provide theoretical and technological support for the assessment of R. esculentum hatchery release. It will also be helpful for future research into R. esculentum species enhancement and germplasm resource conservation.

2 MATERIAL AND METHOD 2.1 Material sampling

The artificial seeding of the R. esculentum were conducted as follows. In the fall, three hundred parental medusa were caught from the wild and then transported to the hatchery, where eggs and sperms were released into the water at the same time. After fertilization, the fertilized eggs began to hatch out and grew into planulae within 8 hours. And then the planulae metamorphosed, settled, and became scyphistomae within 4 days. The scyphistomae underwent asexual reproduction or produced podocysts when the environmental conditions were suitable. They passed through the winter and produced ephyrae by strobilation in the spring. Finally, the ephyrae grew up into young medusa.

A total of 600 R. esculentum medusae averaging 20 mm in diameter were collected from the Qingdao Golden Beach Fisheries Company hatchery. Because of their small size, the entire body of each individual was preserved in 95% ethanol and stored at-20℃ until DNA extracted. Twelve scyphistomae were collected from the hatchery and fed with newly hatched Artemia nauplii once a day in 1-L glass beakers. The released ephyrae from the same scyphistomae were transferred to another 1-L glass beaker, and fed with nauplii twice daily until the ephyrae reached 10 mm in diameter. Each scyphistomae released 2-5 ephyrae comprising a total of 30 individuals that were used for subsequent experiments.

2.2 DNA extraction

Total genomic DNA was extracted from the ethanol preserved tissues, following a standard phenolchloroform protocol (Dawson and Jacobs, 2001). Pure genomic DNA was dissolved in double-distilled water and the concentration was adjusted to 100 ng/μL according to the OD₂₆₀ value, which was measured in a GeneQuant spectrophotometer (Amersham Biosciences) and detected by 1% agarose gel electrophoresis.

2.3 Microsatellite markers and genotyping

The microsatellite sequences used were developed by Zhu et al. (2015). Out of their 15, we selected ten sequences that amplified well and had gene polymorphism in the population. Forward primers were labelled with different fluorescent dyes (6-FAM, ROX, HEX, and TAMRA-Sangon Biotech). The 25-μL PCR master mix contained approximately 100 ng genomic DNA, 10× PCR reaction buffer, 0.5 μmol/L of each forward and reverse primers, 2.5 mmol/L each dNTP, 2.5 mmol/L MgCl₂, and 1 U Taq. Amplifications were performed separately for each locus in an Mastercycler Gradient (Eppendorf Germany) using the following thermal cycling profile: initial denaturation at 94℃ for 5 min followed by 36 cycles of denaturation at 94℃ for 45 s, annealing at the optimum temperature for 30 s, extension at 72℃ for 45 s, and a final extension at 72℃ for 5 min. After the amplification, products were capillary electrophoresed on an ABI 3130XLs using the internal size standard LIZ 500 (Applied Biosystems).

2.4 Data analysis

The chromatograms were analyzed in GeneMapper 4.0 (Applied Biosystems) and the statistics were collated into an Excel^®2013 document. The observed number of alleles (k), observed heterozygosity (H_o), expected heterozygosity (H_e), combined nonexclusion probability, polymorphism information content (PIC), and Hardy-Weinberg Equilibrium (HWE) were calculated in Cervus 3.0 (Kalinowski et al., 2007). Inbreeding coefficient (F_is) was estimated in Genepop 4.3 (Rousset, 2008).

3 RESULT 3.1 Genetic diversity within population

Between 2 (REG-6 and REG-15) and 17 (REG-27) alleles were detected, with a total of 85 (mean=8.5 per locus). Within-population levels of observed (H_o) and expected (H_e) heterozygosity ranged from 0.152 to 0.839 (mean=0.464) and from 0.235 to 0.821 (mean=0.618), respectively (Table 1). The polymorphism information content (PIC) of each marker ranged from 0.207 to 0.795 with an average of 0.580. Of the 10 loci initially selected, seven deviated significantly from HWE (P < 0.001) after standard Bonferroni adjustment.

3.2 Inbreeding coefficient (F_is)

The detailed inbreeding coefficients for each locus are shown in Table 1. Levels of F_is were significantly different from zero and ranged from-0.156 (REG-39) to 0.719 (REG-42), with an average of 0.203.

3.3 Ephyrae genotyping results

When analyzed the genetic diversity using the identity analysis function in the CERVUS 3.0 we found that two medusa shared the same genotype at all ten microsatellite loci. The probability of identity among siblings (P _{(ID) sib}) was 7.88×10^-4. In theory, the ephyrae released should be homogeneous in their genetic information (i.e., clones); we therefore speculated that these two individuals were from the same scyphistomae. In order to prove the hypothesis, thirty individuals were collected and genotyped at the 10 microsatellite loci. As a result, the ephyrae produced by each of the 12 scyphistomae had the same genotype according to the SSR analysis.

4 DISCUSSION 4.1 Genetic diversity within population

The PIC reflected the diversity of loci in the population. A locus is considered highly polymorphic when PIC>0.5. PIC between 0.25 and 0.5 indicates moderate polymorphism and PIC < 0.25 indicates low polymorphism (Botstein et al., 1980). As revealed in Table 1, six of the microsatellite loci were highly informative (PIC>0.5), two were reasonably informative (0.25 < PIC < 0.5), and two were only slightly informative (PIC < 0.25). The average polymorphism content was >0.5, which indicated that these ten microsatellite loci can be applied to effectively analyze the genetic differentiation within a population. In the present study, the observed number of alleles over the ten loci ranged from 2 to 17, which were much higher than the number detected by Zhu et al. (2015) of 2 to 12 for analyzing the genetic diversity of wild R. esculentum based on microsatellite markers. A possible explanation of this larger diversity was that the broodstock might come from several populations, which increased the possibility of detecting more new alleles. In addition, the heterozygosity was also the important indicator of population genetic diversity. The mean observed and expected heterozygosities (0.464, 0.618) we detected were also higher than their findings (0.346 and 0.318), which demonstrated that the hatchery jellyfish population retained a high level of genetic diversity.

The HWE indicates that gene and genotype frequency remain at a stable level in a population from generation to generation in the absence of other evolutionary influences. In the present study, eight of the 10 selected loci deviated significantly from HWE after Bonferroni correction. These results were similar to those of Aglieri et al. (2014). They found that eight out of nine loci deviated significantly from HWE when they analyzed a population of the holoplanktonic Jellyfish Pelagia noctiluca (Scyphozoa: Cnidaria) from eight areas in Italy. The average observed heterozygosity (0.464) in this study was less than expected (0.617), which indicated a heterozygote deficit in the population and also resulted in HardyWeinberg disequilibrium (HWD). HWD resulted from a heterozygote deficiency at all of the loci, there are two possible explanations: either the presence of technical artefacts such as null alleles or the influence of biological factors (Callen et al., 1993). Moreover, sample size in an experiment may also affected the HWE (Yu and Chu, 2006). R. esculentum breeding technology was not mature and the parents were obtained by catching the wild individuals. Genetic drift and bottlenecking could have easily occurred in the population because of the limited numbers of parents, which resulted in the changes of allele frequency (Sun et al., 2010). Furthermore, R. esculentum has a complex life history that includes both a sexual and an asexual process. The fact that clones from several scyphistomae are found would be an additional potential explanation from HWE deviations. And the asexual process is easily affected by biological factors, including nutrition, temperature, and light (Lu et al., 1997; Zhao et al., 2006; You et al., 2010; Wang et al., 2014b). These environmental factors would also have brought about changes in allele frequency in the jellyfish population to a certain extent. Based on this, we conclude that a heterozygote deficiency and biological factors eventually led to the deviation from HWE at multiple loci.

4.2 Inbreeding coefficient (F_is)

The inbreeding coefficient (F_is) ranges from-1 to 1 and the population showed a certain degree of inbreeding when the F_is value was highly significantly positive. As shown in Table 1, the inbreeding coefficient at four loci was < 0 and >0 at the other six loci with an average of 0.203, indicating a certain degree of inbreeding in the hatchery population. There is no unified theory to accurately explain inbreeding levels for aquatic animals at present. Moss et al. (2007) suggested that the inbreeding value should controlled to within 0.1 to avoid inbreeding effects on survival and growth. Inbreeding depression has become a common problem in cultured aquatic animals. However, there have been no reports on inbreeding in R. esculentum. In view of its complex life cycle, inbreeding is more likely to occur within R. esculentum populations than in other aquatic animals. Although, the parents used for breeding were wild caught and the numbers were low, R. esculentum exhibits external fertilization and high fecundity. Broadcast spawning would theoretically ensure homogeneous distribution of gametes and random fertilization in the water column and then behavioural features may drive patterns of kin aggregation. Moreover, asexual reproduction increases the number of homogeneous individuals, which further generate more half and full siblings. Over time, serious inbreeding depression would occur in the subsequent generations. Inbreeding depression can influence many traits. In general, the traits governing environmental adaptation are more likely to show inbreeding depression than morphological traits, such as viability, reproductive ability, and competition ability (Ma et al., 2005). This would have a serious effect on wild populations if this R. esculentum population was released into the sea. Therefore, our findings may be helpful in selecting groups for release and maintaining genetic diversity in populations.

4.3 Estimating the probability of identity among R. esculentum clones

The R. esculentum life history includes both sexual and asexual reproduction. Strobilation (or transverse fission) through the spontaneous transverse segmentation of the body has been defined as a type of asexual reproduction. Therefore, in theory, the ephyrae released should be homogeneous in their genetic information and be the clones (Qiao et al., 2013). In the present study, two medusa genotypes were identical at ten microsatellite loci and the probability of identity among sibs (P _{(ID) sib}) was 7.88×10^-4. The individual probability of identity is the probability, given the genotype of one individual, that a second individual will have the same genotype, assuming no typing errors occur. The P _{(ID) sib} value is related to the number of microsatellite loci and their polymorphism. Generally, low P _{(ID) sib} values are desirable (Eichmann et al., 2005). Waits et al. (2001) suggested a P _{(ID) sib} between 0.001 and 0.000 1 as sufficiently low for the identification of individuals in animal populations. Our P _{(ID) sib}value (7.88×10^-4) was between 0.001 and 0.000 1, which illustrated that the 10 microsatellite loci used can effectively identify the clones. Additionally, we separated several polyps and each one eventually split into 2-5 ephyrae. The ephyrae were genotyped at the ten microsatellite loci and the results revealed that the genotypes of individuals from the same polyp matched exactly. Based on the above, we conclude that these two individuals were derived from the same polyp and had the same genetic material.

Accurate assessment of the recapture rate is an urgent problem for jellyfish releasing research. Our research revealed that the probabilities of exclusion based on the genotype of no parent known and one parent known were 97.26% and 99.80%, respectively. When both parents were known, the average exclusion probability was 99.99%. Therefore, the 10 microsatellite markers we used can not only be used to analyze the identity of individuals but can also be applied to parentage identification. The results provide a theoretical basis and technical support future studies to assess R. esculentum enhancement based on molecular markers. Our study also promotes the use of research on hatchery R. esculentum prior to release.

5 CONCLUSION

The hatchery population in this study has maintained a high level of genetic diversity, while a certain amount of inbreeding has occurred in the population. We did not find any variation among the ephyrae from the same scyphistomae at the clone level using the microsatellite marker analysis. We believe that releasing this R. esculentum group into the wild will improve the genetic diversity of germplasm resources. We should optimize the population structure of jellyfish to avoid inbreeding depression in subsequent generations. Microsatellite molecular marker technology used in other aquatic animal enhancement programs can also be applied to R. esculentum hatchery releases. This research provides a theoretical basis and technical support for genetic diversity detection in and selection of R. esculentum populations. These findings support the use of releasing studies and germplasm resource conservation in R. esculentum.