Modeling the impact of plant structure on host-finding behavior of parasitoids.

For many parasitoid species, the final step of host location occurs on plants whose structure varies in time and space, altering the capacity of parasitoids to exploit hosts. Plant structure can be defined by its size, heterogeneity and connectivity. We tested the hypothesis that these three components and all possible interactions affect the level of parasitism of Trichogramma evanescens and that parasitism can be predicted if the structure of a plant is measured. We quantified and varied the structure of three-dimensional artificial plants to determine which component(s) of plant structure explain variability of parasitism and to develop a model that predicts parasitism by Trichogramma females. This model was validated with three natural tritrophic systems. The experiment with artificial plants revealed that plant structure affected host-finding success, which was higher on plants with a simple structure and low on plants with a complex structure. A response surface regression showed that the linear and quadratic terms of connectivity were highly significant, indicating that connectivity best explained the variability in the rate of parasitism obtained. The interaction between connectivity and heterogeneity was also significant. Observed values of parasitism from experiments with three natural tritrophic systems fit predicted values of parasitism generated by the model, indicating that parasitism can be predicted if heterogeneity and connectivity of a plant are known. Consequences of these results in regard to population dynamics, evolution and biological control are discussed.