EVOLUTION OF CONDITION-DEPENDENT DISPERSAL - A GENETIC-ALGORITHM SEARCH FOR THE ESS REACTION NORM

Many insects produce two types twinged and wingless) of offspring that greatly differ in dispersal ability. The ratio of the two types often depends on the quality of the local habitat and the crowding experien ced by the mother. Here we studied the condition-dependent dispersal t hat is evolutionarily stable. The model is also applicable to annual p lants that produce two types of seeds differing in dispersal rates. Th e model assumptions are: the population is composed of a number of sit es each occupied by a single adult. The total number of offspring prod uced by a mother depends on the environmental quality of the site that varies over the years and between sites. The ESS fraction of dispersi ng type as a function of the quality of the habitat (or ESS reaction n orm) states that dispersers should not be produced if habitat quality m is smaller than a critical value k. If m is larger than k, the numbe r of dispersers should increase with m and that of nondispersers shoul d be kept constant. Second, we developed an alternative way of searchi ng for the ESS: the reaction norm is represented as a three-layered ne ural network, and the parameters (weights and biases) are chosen by ge netic algorithm (GA). This method can be extended easily to the cases of multiple environmental factors. There was an optimal (relatively wi de) range of mutation rates for weights and biases, outside of which t he convergence of the network to the valid ESS was likely to fail. Rec ombination, or crossing-over, was not effective in improving the succe ss rate. The learned network often shows several characteristic ways o f deviation from the ESS. We also examined the case in which the quali ty of different sites was correlated. In this case the ESS fraction of dispersers increases both with the quality of the site and with the a verage quality of the whole population in that year.