Ulations of Drosophila Pseudoobscura

Most natural populations of Drosophila pseudoobscura are mixtures of individuals homozygous and heterozygous for gene arrangements in the third chromosome. The relative frequencies of the gene arrangements in populations show geographical and temporal variations. On Mount San Jacinto and in the Sierra Kevada in California, and in some localities in Texas, the temporal variations are cyclic and connected with the succession of the seasons (DOBZHANSKY 1943, 1948, DOBZHANSKY and EPLING 1944). I t has been inferred that the seasonal changes in the frequencies of the gene arrangements are caused by natural selection. Carriers of different chromosomal types possess different adaptive values, and different types are favored a t different seasons. This hypothesis has been indirectly confirmed by experiments on artificial populations which contained mixtures of different gene arrangements. Under certain conditions, changes in the relative frequencies of gene arrangements are observed from generation to generation in such artificial populations (WRIGHT and DOBZHANSKY 1946, and unpublished data). Analysis of these changes shows that, with rare exceptions, individuals of which the third chromosomes differ in gene arrangement (inversion heterozygotes) are superior in fitness to inversion homozygotes. An independent proof of the adaptive superiority of the heterozygotes to the homozygotes is as follows: Among the eggs deposited in the cages containing the artificial populations, inversion heterozygotes and homozygotes occur with relative frequencies approaching that demanded by the HARDY-WEINBERG rule. Yet, among the adult flies developed in the same population cages, excesses of heterozygotes and deficiencies of homozygotes are observed (DOBZHANSKY 1947a). I t follows, then, that carriers of different gene arrangements interbreed a t random in the population cages, and that a differential mortality between the egg and the adult stage upsets the HARDYWEINBERG equilibrium proportions. As a working hypothesis, it may be assumed that in natural populations likewise, the fitness of inversion heterozygotes is higher than that of homozygotes, and that the observed temporal changes in the natural populations are connected with this difference in fitness. If this hypothesis is valid, the homozygotes will be less and the heterozygotes will be more frequent among the flies found in nature than would be expected on the basis of the HARDY-WEINBERG equilibrium. The present article describes an attempt to test this working hypothesis.