Geometry for herbivore colonization of crop fields

Background/Question/Methods
There is a pervasive assumption in agricultural ecology that fine-grain spatial heterogeneity is better for pest control than course-grain heterogeneity. It seems intuitive to many observers that having a larger number of smaller fields of a crop leads to lower pest densities than having fewer large fields, and this expectation has been proposed to contribute to the worsening of pest problems associated with agricultural intensification. However, there are both empirical and theoretical results that suggest that the conventional wisdom may be incorrect in some cases. For example, recent studies of the Andean potato weevil (Premnotrypes spp) have demonstrated that it is less  damaging in agricultural landscapes that had fewer, larger potato fields. We developed a spatially explicit simulation model of insect movement across an agricultural landscape, motivated by the weevil study system, to uncover basic processes that could lessen pest impact in larger crop fields. Simulations, implemented in MATLAB, included stochastic field placement and deterministic herbivore behavior and movement, and emphasized processes by which herbivores colonize new crop plantings from overwintering sites. Crop size was manipulated by clustering together sets of smaller fields; thus, field size was manipulated without changing the total proportion of the landscape devoted to the focal crop type.

Results/Conclusions
Our results provide plausible mechanistic explanations for the results of the previous field studies. The models show that in this system field aggregation led to severely reduced pest densities, with weevil populations declining to extinction in only a few years when crop fields had been highly aggregated to create large monocultures. The herbivore population growth rate was negatively impacted by the increased average distance between overwintering sites and new crop fields as fields were clumped together. Additionally, herbivores with no ability to sense crops from a distance, and which therefore dispersed at random and colonized crops only upon entering fields, had a lower probability of encountering a smaller number of larger fields compared to a larger number of smaller fields. The features that drove the negative relationship between aggregation and herbivore densities include short dispersal distances, no alternate host plants, and no predation. These results offer a novel approach to managing the Andean potato weevil, and can be generalized to build an intuition for when field aggregation may lead to reduced herbivory in agroecosystems.