Alternative models for allozyme-associated heterosis in the marine bivalve Spisula ovalis

Correlations between allozyme heterozygosity and fitness-related traits, especially growth, have been documented in natural populations of marine bivalves. However, no consistent pattern has been exhibited, because heterotic effects on size vary with age and individual growth parameters are generally unknown. No consensus has emerged on the genetic basis of allozyme-associated heterosis. The species studied here, Spisula ovalis, displays annual shell growth lines, which allows us to compute individual age and growth dynamics over the whole life span. Our morphological study was coupled to a protein electrophoresis study at seven polymorphic loci. While the maximum size gained is not related to heterozygosity, the age at half maximum size, t1/2, is significantly negatively correlated with heterozygosity, indicating an heterotic effect on initial growth. The correlation between heterozygosity and size is expected to vanish when age increases, due to the form of the growth function. This decreasing correlation is consistent with previous studies. We compare the relative performances of five linear models to analyze the genetic basis of heterosis. Surprisingly, the largest part of variance in t1/2 is due to additive effects, the overdominant components being much weaker. Heterosis is therefore due to general genomic effects rather than to local overdominance restricted to allozymes or small neighboring chromosomal segments. A significant dependence of individual heterotic contributions of the enzyme loci upon expected heterozygosities, rather than metabolic function, further supports the hypothesis of enzymes acting as markers. General genomic effects can hold only if allozyme heterozygosity is positively correlated with heterozygosity at fitness-related genes scattered throughout the genome. This hypothesis is supported here by heterozygosity correlations between enzymatic loci.