Abstract

The shortcomings of present population genetic theory are discussed as they pertain to problems of speciation, extinction and the evolution of genetic systems. It is suggested that the modern theory of games may be useful in finding exact answers to problems of evolution not covered by the theory of population genetics. An outline of relevant topics in the theory of games is given. It is suggested that the most pertinent utility measure for a population is its one-generation probability of survival and that a strategy or a mixture of strategies corresponding to a maximin strategy will be found in natural populations. These notions are applied to a population segregating for two alleles with different norms of reaction in different environments. For the model chosen the optimal strategy is found to be homozygosis for different alleles in different populations due either to inbreeding or genetic isolation. A segregating polymorphism in such populations would be a detriment to the species, although the heterozygotes are more constant in fitness.