In order to assess the relationship between genetic and environmental variability, a large natural population of Drosophila melanogaster was replicated as eight subpopulations, which were subjected to four different patterns of environmental variation. The environmental variable imposed was presence of 15% ethanol in the culture medium. Experimental treatments of the populations were intended to simulate constant environmental conditions, spatial heterogeneity in the environment, and two patterns of temporal environmental variation with different periodicity (long- and short-term temporal variation). Additive genetic and phenotypic variation in sternopleural and abdominal chaeta number, and body weight, were estimated in two successive years, and measurements were taken of the genotype–environment correlation of body weight and sternopleural bristle score with medium type.
Additive genetic variance of sternopleural chaeta number and of body weight was significantly greater in the three populations experiencing environmental heterogeneity than in the control population, but additive genetic variance of abdominal bristle score was not clearly affected by exposing populations to varying environments. Temporal environmental variation was equally, if not more, efficient in promoting the maintenance of genetic variation than spatial heterogeneity, but the cycle length of the temporal variation was of no consequence. Specific genotype–environment interactions were not present, therefore adaptation to heterogeneous environments is by selection of heterozygosity per se, rather than by differential survival of genotypes in the alternate niches.