GENOTYPE-ENVIRONMENT INTERACTION FOR CLIMATE AND COMPETITION IN A NATURAL POPULATION OF FLOUR BEETLES, TRIBOLIUM CASTANEUM.

The norm of reaction, the set of average phenotypes produced by a genotype in different environments, can be affected by spatial variation in natural selection especially when there exists genotype-environment interaction. In subdivided populations, the greater the genotype-environment interaction variance and the lower the migration rate, the more independent are the possible evolutionary trajectories for local adaptation. I examined genotype-environment interaction in the rate of population increase for lineages randomly derived from a wild population of Tribolium castaneum across a series of ecologically important environments. The lineages were derived from an outbred, wild-caught population by 14 generations of random genetic drift, during which the effective size of each lineage was approximately 22 breeding adults. The environments studied were the classic temperate-wet and cold-dry climates of Park (1954) in factorial combination with two genetic strains of a congeneric competitor, T. confusum. Much among-lineage genetic variation for rate of population increase was found for each of these ecologically important environments of climate and competition. Genotype-environment interaction accounted for 40.5% of the total among-lineage variance in rate of population increase signifying that the performance of a lineage in one environment is not necessarily a good predictor of its performance in another. Changing the genetic identity of the competitor changed the rate of increase of some lineages as much or more than changing the climatic conditions of temperature and humidity. This is the first empirical study to characterize the genotype-environment interaction variance associated with genetic variation in a competing congeneric species. This competitor-specific genetic variation in competitive ability may play an important role in coevolution in subdivided populations.