Genetic diversity assessment of Pacific oyster Magallana gigas (Crassostrea gigas) populations from the two southern coastal farms in South Korea

South Korea is among the major producers of the Pacific oyster, Magallana gigas (Crassostrea gigas), which is one of the most valued aquaculture species. Since the early 1990s, climatic and anthropogenic factors have incurred the reduction of their wild seeds, whereby the dependence on hatchery-produced seeds has constantly increased in South Korea, thus raising concerns about losing genetic diversity and accelerating genetic deterioration. To better understand their genetic make-up, we assessed the genetic diversity of M. gigas populations from two farms (Tongyeong and Gadeokdo) in the southern coast, where about 80% of the cultivated oysters in Korea are produced. Tongyeong showed slightly higher diversity than Gadeokdo, but both populations had a similar genetic structure characterized by low nucleotide diversity. Comparative haplotype analyses provided data supporting unique genetic features of the populations that include (1) weak genotype-locality relationship, (2) low levels of gene flow between populations, and (3) seasonal fluctuation of genetic variation within a population. Furthermore, the highly alike haplotype network patterns were observed between the wild and farm populations as well as among the populations in neighboring countries, which suggests that the genetic structure is conserved between wild and hatchery populations, and geographic proximity has minimal influence on the genetic composition. These results warrant further study in biological and ecological contexts and will be invaluable in formulating genetic monitoring and sustainable long-term management of M. gigas.


39
The Pacific oyster, Magallana gigas, formerly Crassostrea gigas, is a seasonal delicacy beloved all over of its small size, maternal and non-recombining mode of inheritance, rapid evolution, extensive intraspecific polymorphisms, etc [15,16]. Among mtDNA markers, cytochrome oxidase c subunit I (mtCOI) is one of the 66 most conserved protein-encoding genes in the mitochondrial genome [17]. The effectiveness of mtCOI analysis 67 to study genetic diversity and genetic structure of marine mollusk populations has been charted in Tegillarca  of each pairwise comparison was tested with 10,000 random replicates.  and 0.000689, respectively), with TY being slightly more diverse (Table 2). Consistently, the haplotype 160 diversity of TY was higher than that of GD. The mean number of pairwise differences (k) measuring the genetic 161 diversity within a population was higher in TY than in GD (  Table 3, a majority of the total molecular variance was present within populations (99.81%) rather 172 than between the 2 populations. Accordingly, the fixation index (FST), a measure of population differentiation 173 due to genetic structure, was 0.00188 but was not statistically significant (p-value = 0.49853).

190
In addition, the seasonal pattern of haplotypes was accessed in TY and GD populations ( HAPs were included in this analysis (see Table 1). Similar to Fig 3,  haplotypes shared by different geographical origins were identified (Fig 4). Despite the geographical proximity,

Genetic features of M. gigas populations 227
A total of 21 haplotypes were identified in TY and GD populations, based on SNPs within the mtCOI gene.

228
The subsequent genealogical analysis showed that these haplotypes form a characteristic shallow network 229 consisting of the major haplotype (i.e., Hap_1) at the center and 20 minor satellite haplotypes diverging from 230 HAP_1 (Fig. 3). Genetic differences between the two populations were shown to lie in their distinctive satellite 231 haplotypes; TY had 12 satellite haplotypes while GD had 8, but none of the satellite haplotypes were shared 232 between the two. Each satellite haplotype differed from HAP_1 by 1-3 mtCOI SNPs and mostly formed a 251 gigas populations. Therefore, it is plausible that gene flow between TY and GD populations is not significant given the ocean ecology of these two regions is believed to be commonly affected by the Kuroshio current [34].

254
The finding that TY had slightly higher variation than GD suggests that fine-scale environmental factors might 255 play a role in the equilibrium between gene flow and genetic drift, which needs further study [35].

256
As the samples from TY and GD were collected over the four seasons, we examined temporal correlation 257 between season and haplotype diversity. Distinct haplotypes constituted each season, and frequent replacements 258 of haplotypes were detected over the 4 seasonal periods except for Hap_1 (Table 4). This data suggests that 259 genetic variation was not passed down from one season to the next (i.e., the absence of self-recruiting). This

277
Information from such study will be the groundwork to set strategies for reducing genetic erosion and 278 developing more sustainable management and breeding program for M. gigas populations.