Experimental Study of Speciation in Ecological Niche Theory Using Genetic Algorithms.

Abstract

This thesis is concerned with using genetic algorithms to investigate ecological niche concepts. A Genetic algorithm is an adaptive search method based on models from mathematical population genetics. It generates adaptive responses by simulating the information processing achieved in natural systems by means of the mechanisms of heredity and evolution. This is accomplished internally by maintaining populations of individuals (species). In this simulation research the (individuals) structures of the genetic adaptive algorithm will be given species tags which are generated at r and om and have no direct effect upon fitness. They can be considered incidental phenotypic markers that define and enforce the choice of mates for these structures. The research question then becomes: Will there develop an association between species tags and some of the schemata conferring fitness in specific niches? That is, will we find that given species become associated with given niches despite the fact that the species tags themselves are chosen r and omly and do not directly confer fitness? If this is so, and we will see that it is, then a case can be made for the selection of incidental phenotypic properties as a means of providing non-uniform mate selection to the point of eventually giving rise to distinct species. Because the isolation is the result of internal processes rather than environmentally imposed conditions (such a geographic isolation), the result is an interesting case of speciation under panmictic conditions.