Genetic variability of an orange-shell line of the Pacific oyster Crassostrea gigas during artificial selection inferred from microsatellites and mitochondrial COI sequences

Rare breeds represent a valuable resource for current or future market demands, but their low census compromises the genetic variability and future utilization of these breeds. Whether genetic variability of rare breeds with low initial genetic variation can be maintained during an intense mass selection becomes the key to applying mass selection to the genetic improvement of these breeds. The genetic variability among three generations of successive mass selection of the orange-shell line of Crassostrea gigas (MS1–MS3) were evaluated by both 20 nuclear microsatellite loci and mitochondrial cytochrome oxidase I sequences (mtCOI) compared to four wild populations. In this study, the orange-shell line exhibited very low genetic variability. Only one mtCOI haplotype was detected in all individuals of MS1–MS3. Significant reductions in average number of alleles (Nₐ: 69.55–76.92%), allelic richness (Aᵣ: 68.17–74.91%) and expected heterozygosity (Hₑ: 34.21–39.24%) as well as increased mean pairwise genetic relatedness (R: 6.87–25.79 times) were observed in MS1–MS3 when compared to wild populations. However, the genetic variability of orange-shell line with very low initial variation successfully maintained during three generations of mass selection. No significant difference in Nₐ (3.60–4.40), Aᵣ (3.51–4.08) and Hₑ (0.48–0.50) occurred among MS1–MS3. Pedigree reconstructions (no full-sib group from MS1–MS3 was larger than 16% of the whole group) revealed artificial spawning used in this study has a better control over contribution of gametes than traditional mass spawning. In addition, effective population size of MS1–MS3 calculated by linkage disequilibrium methods increased from 29.3 to 67.0 indicating the linkage disequilibrium decays over time. This study provides important insights in the genetic consequences of a rare variant line of C. gigas with very low genetic variation over generations of mass selection. This will provide a reference for carrying out genetic improvement programs on rare breeds where small populations are inevitable.