The Relation between Expressivity and Selection against Eyeless in Drosophila Melanogaster.

ANY mutant stocks of Drosophila mass cultured in the laboratory for generaM tions become phenotypically almost indistinguishable from wild type. Upon outcrossing, the mutant reappears in the FZ with its pristine strength of expression. Presumably, genes reducing the degree of expression, or expressivity, of the major character of the mutant stock had appeared and accumulated through natural selection during its maintenance in the laboratory. These modifying genes are probably disadvantageous to their possessors in a culture of wild type flies. MARSHALL and MULLER (1917) found more extreme phenotypes in homozygous recessive mutants obtained from stocks maintained in a heterozygous condition for many generations, than in parallel stocks maintained as homozygotes. Does lower expressivity reduce the rate of selection against a deleterious gene when compared with its wild type allele? If so, is the relation demonstrable within a phenotypically variable isogenic stock or must genetic modifiers exist? The rate of selection is determined by the viability and productivity of flies of the relevant genotypes. The first of these parameters can be measured in crosses giving equal numbers of mutant and wild type zygotes whereas the latter parameter can be estimated from experiments in which mutant and wild type males compete in mating, or from experiments in which mutant and wild type females compete in oviposition. The effect of the degree of expression of the mutant on viability and productivityand thus on selection-can be studied in crosses involving mutant parents with different grades of expression. The next question is whether any correlation observed between expressivity and selection is to be attributed to the accumulation of modifiers. If the degree of expression is influenced by modifiers, there should be a positive correlation between the expression of the mutant character in parent and offspring. Furthermore, if the character is bilateral, the grades on the two sides are correlated to a greater extent within the progeny of a group of parents heterozygous for modifiers than of homozygous parents. The reason for this is that with heterozygous parents both local environment and genetic segregation affect the individual offspring whereas with homozygous parents only the local environment determines differences in expression. The results to be reported demonstrate that modifiers soon disrupted the initial isogenicity of the mutant stock, enhancing its viability and productivity and decreasing its expressivity, so long as the mutant stock was maintained separately from the wild type stock. Once mutant and wild type parents were taken from segregating cultures, the accumulation of modifiers was greatly retarded. Neverthe-