PREMATING ISOLATION IN A SIMULATION-MODEL GENERATES FREQUENCY-DEPENDENT SELECTION AND ALTERS ESTABLISHMENT RATES OF RESISTANT NATURAL ENEMIES

In a stochastic simulation model, rates of establishment of new strains of arthropod natural enemies were examined under different assumptions for the following four factors: (1) mating biases, (2) diploidy state (diplo- or haplo-diploid), (3) degree of dominance of the resistance allele, and (4) degree to which mating biases extend to heterozygotes. In the simulation results, the amount of symmetrical mating bias (preference of females to mate with males of the same genotype) was an important factor in determining the rate of establishment when released individuals initially comprised 10% of the population. All interactions between mating bias and the other three factors were also significant. Diploidy level had no main effect but strongly interacted with other factors. When mating was random in simulations of haplo-diploid populations, the resistance allele was established faster than simulations of diplo-diploid populations. In those populations with no interbreeding between strains, however, diploidy level had no effect on the rate of establishment of the resistance allele. As we expected from previous resistance models, the dominance of the resistance allele significantly affected the rate of resistance development. When dominance varied, rates of resistance development for diplo-diploid simulations varied more than with corresponding haplo-diploid simulations. The degree to which mating biases extended to heterozygotes depended on the levels of the other factors, especially the diploidy level. Our simulation results suggest that mating biases can have a large effect on the establishment of strains of arthropod natural enemies and in some cases may be capable of preventing establishment of strains that are better adapted to local environments. If our simulations accurately represent pest management programs, incomplete mating biases between the strain to be released and strains previously established might reduce efficacy when releases of small numbers of individuals are made.