Investigation on the founder effect

The fitness of threeDrosophila melanogaster populations (A,B,C) and the crosses between them (A×B; A×C; B×C) has been studied in conditions of overlapping generations. Each line was subdivided into 4 sub-populations (1,2,3,4) differing in the phenotype of the founder flies (vestigial or wild type) and of which sub-populations 2 and 3 had the same gene composition. Fitness was estimated as the number of adult flies present in each population. The results suggest that: (1) a differentiation of fitness values is detectable between sub-populations founded from groups of flies having the same gene composition; (2) isolation does not seem to be a necessary condition for the observed differentiation. These results are discussed in connection with the evolutionary effects of random genetic drift. A working hypothesis has been advanced which would explain the observed results as dependent on complex interactions involving sexual dimorphism and interactions between genes.