Hybrid vigor in Pacific oysters: An experimental approach using crosses among inbred lines

Two competing genetic hypotheses for heterosis, dominance and overdominance, have been championed to explain positive correlations between allozyme heterozygosity and fitness-related traits for bivalve molluscs. To begin to test these hypotheses, we made controlled crosses among inbred lines of the Pacific oyster Crassostrea gigas. In such mating experiments, heterosis (h(p)) can be defined and quantified through ANOVA as Q/L > 1.0 or < - 1.0, where L is the trait-difference between the two parental inbred lines and Q is twice the deviation of the hybrid from the mid-parent value (Griffing, 1990). Inbred lines of the Pacific oyster were made by selfing hermaphrodites; brood stock pedigrees and a grand mean fixation index of F = 0.5 were confirmed by allozymes. In two separate experiments pairs of inbred lines were crossed in 2 X 2 fashion to produce two inbred and two hybrid progeny genotypes. Each genotype was replicated by pairwise matings and replicates were grown in multiple containers (81 plastic bags), from which data on larval mortality (proportional decrement in population number per day) and size (shell-height);were obtained. In the first experiment, the two inbred genotypes differed significantly from each other in daily larval mortality ( -0.115 vs. -0.881) and both hybrids had lower mortalities than inbred genotypes (-0.034 and -0.078); mean h(p), 1.15, was significantly greater than 1.0. By Day 14, larvae of one hybrid group were significantly smaller than those of the inbred genotypes (241.0 mu m vs. 253.5 and 259.2 mu m, respectively); only a very small part of this difference could be attributed to negative correlation between larval density and growth. Negative heterosis for larval size was highly significant (h(p) = -5.36, P<0.001) and persisted to the juvenile stage (mean h(p) = -1.34, ns, for shell-height at Day 154 and h(p) = -7.77, P<0.001, for live-weight at Day 163). In the second experiment, inbred genotypes again differed significantly in daily larval mortality ( -0.341 vs. -0.654). One hybrid group had significantly less daily mortality than either parent ( -0.110), but the other was similar to the best parent (-0.347), so that for one hybrid, h(p) = 2.47, P<0.01, and for the other, h(p) = 0.96, ns). Variance among genotypes for larval shell-length was highly significant on Days 2, 7, and 14, with heterosis evident on Days 2 and 7 (h(p) = 1,04 and h, = 3.84, respectively). At 11 months of age, hybrid shell-height averaged 150% of the shell-height for the best inbred parent and hp approximate to 7.7. These measurements of heterosis, both positive-and negative, implicate a third explanation of heterosis, epistasis. Intercrosses of F-1 hybrids can be used to discriminate among dominance, overdominance and epistasis hypotheses for quantitative trait loci (QTL) causing hybrid vigor.