INFERENCES ON THE NATURE OF GENETIC VARIATION FOR VIABILITY FROM ARTIFICIAL SELECTION AND INBREEDING IN DROSOPHILA

INTRODUCTION The genetic properties of life-history traits have been investigated for a long time. On the one hand, a number of estimates of heritability and inbreeding depression are available for reproductive traits in a variety of species (Lynch and Walsh, 1998). On the other, estimates of mutational parameters, such as mutational variances and rates and effects of deleterious mutations, are also available for such traits (Lynch et al., 1999 ; Garcia-Dorado et al., 1999). These two sorts of information can be combined, assuming a simple genetic model, to obtain information on the genetic architecture of life-history traits. We carried out an experiment of artificial selection and inbreeding with a natural population of Drosophila melanogaster, and obtained estimates of selection response and inbreeding depression for egg-to-adult viability. Using a range of mutational parameters and assuming a simple deleterious mutation-selection balance model of variation, we simulated the experimental conditions and compared the simulated results with the empirical observations. Our results suggest that a model assuming mutation-selection balance can explain the frequency of genes affecting viability in the natural population, but contrasting models of mutation cannot be discarded.