Studies of the Relative Fitnesses of Experimental Populations of Drosophila melanogaster

Abstract

This paper deals with an attempt to estimate the relative fitnesses of several experimental populations of D. melanogaster. The measure of fitness used was the number of offspring produced by single pairs of individuals. Each experiment consisted of a study of two "competing" populations. Fifty males and 50 virgin females from each of two populations were mated as single pairs in vials. The number of sterile matings was recorded; the offspring produced by fertile pairs were counted fourteen days after mating. Males and virgin females were collected from these vials to establish a second generation of 50 single pairs in vials. The experiments were continued in this way for ten generations. All possible pairs of eight populations (see table 1) were studied. Among the flies representing populations 1 and 3, frequencies of sterile matings increased in successive generations to a high level; this increase must be ascribed to selection of sterility genes in heterozygous condition. In contrast with the increase observed in populations 1 and 3, the frequency of sterile matings decreased in populations 5, 6 and 7. In the remaining three populations (Nos. 17, 18 and 19) the frequency of sterile matings was constant at a relatively low level. Two of the experimental populations (Nos. 1 and 7) produced more offspring per fertile mating in the first generation than did the control population (No. 3). However, every large, irradiated population suffered a progressive decline in the relative numbers of offspring produced per fertile mating in successive generations under the conditions of these studies. Three populations (Nos. 5, 17 and 18) did not show this decline; these either are small populations or have been derived from a small population. This is interpreted as the effect of an abrupt change in population size on fitness. It is also regarded as further evidence for the retention of genes in large populations on the basis of their action in heterozygous condition. Estimates of fitness varied among different samples of flies taken from the same population, among different generations of the same sample, between the two components under observation (fertility and number of offspring produced), and at different times during the course of these experiments. The fitness of a population is not a constant which can be measured, recorded, and then referred to when occasion demands. These results and the difficulties encountered in interpreting such information are discussed.