ABSTRACT
Recently, the gene-for-gene host-parasite coevolution model of Leonard was extended by incorporating two kinds of perturbations. The first kind was the natural perturbations that include those caused by pathogen migration between the two subpopulations of the host, forward and backward mutations in the host or pathogen populations, and some others. The second kind was human perturbations, such as constantly increasing the percentage of the resistant genotype within the host population each season. In this study, we quantitatively compared the two kinds of perturbations and extended the constantly changing human perturbation to include non-constant perturbations that are more likely to occur in the real world. Two properties of the modified Leonard model were revealed from this study. First, when both human perturbations and natural perturbations are involved, the effects of natural perturbations are very small compared with those of human perturbations. This finding ensures that, in the study of human perturbations, we can simplify the study by ignoring the effects of natural perturbations. Second, through the simulation of nonconstant perturbations, which assumes that the proportion of the resistant genotype of the host population increases over time, we found that the model reproduces the “boom and bust” epidemic cycles that are often found in agroecosystems.
