Heritability Estimates For Growth-related Traits And Establishment Of Growth Model In The Pacific Oyster(Crassostrea Gigas)

The study includes two parts. In the first part, a full-factorial mating design consisting of9sires and5dams was employed to estimate heritability for growth-related traits in12-monthPacific oyster, Crassostrea gigas using microsatellite markers for parentage assignment. Inthe second part, the development pattern of C.gigas at different growth stages wasinvestigated by model fitting with regularly collected biological data. The major results andconclusions are as follows:1. Parentage assignment and genetic diversity analysisA total of335sampled progeny were genotyped using six microsatellite markers（ucdCg129, ucdCg134, ucdCg148, ucdCg149, ucdCg160and ucdCg196）:270individualswere successfully assigned to a single parent, while the rest65were identified ascontaminants.Average number of alleles in the offspring and broodstock were both12.67, whichindicated that there was no allele missing in the offspring. The average of observedheterozygosity （Ho） and expected heterozygosity （He） of the broodstock were0.714and0.918respectively, while in the offspring the average were0.916and0.881respectively.Wilcoxon matched pairs test indicated that compared with the broodstock, the Ho of theoffspring was slightly higher （P=0.057）, while the He was significantly lower （P=0.004）.2. Genetic parameters for target traits of the Pacific oysterHeritabilities of shell height, shell length, shell width and wet weight were0.49±0.25,0.36±0.19,0.45±0.23and0.35±0.172, respectively. These estimates falling in the0.3–0.5range belonged to medium-high heritability. Genetic and phenotypic correlations between thefour traits were all positive, with medium-high values （0.55to0.86） for genetic correlationsand low-medium values （0.31to0.50） for phenotypic correlations. Genetic correlationsbetween wet weight and other growth-related traits in descending order were shell length> shell height> shell width, while the phenotypic correlations were shell length> shell width>shell height.3. The growth model of C.gigas at different growth stagesAt larva stage, the relationships between shell height （H）, shell length （L） and age （t） bothconformed to Logistic model, and the growth equations respectively were H=455.612/(1+9.500e^-0.142t), R²=0.999; L=462.476/(1+8.026e^-0.108t), R²=0.996. At larva stage, shellheight and shell length was linearly related, and the regression equation was L=0.76H+18.82, R²=0.994. The growth pattern of C.gigas during grow-out showed a significantseasonal variation. The Polynomial regression equations between shell height （H）, shell length（L）, shell width （W）, total weight （TW） and month age （X） were H=-0.0297X⁴+1.0365X³-12.0220X²+57.6500X-68.9260, R²=0.985; L=-0.0173X⁴+0.5893X³-6.5702X²+30.2420X-34.4150, R²=0.986; W=-0.0068X⁴+0.2620X³-3.2806X²+16.9170X-22.1410, R²=0.956; TW=-0.0219X⁴+0.8234X³-10.1680X²+50.7040X-85.4110, R²=0.972, respectively. Shell height, shell length and shell width all showed significant powerfunction correlativity with total weight. The regression equations respectively were H=23.645TW^0.3213, R²=0.998; L=12.337TW^0.3776, R²=0.995; W=6.611TW^0.3589, R²=0.981.