Models for integrated pest management with chemicals in atmospheric surface layers

Abstract

Surface layer landscapes are described by spatial distributions of canopy type and roughness, surface slopes and topography. Agricultural landscapes also include distributions of pests and pest management systems. Insect populations together with pheromone traps or dispensers constitute a system that depends on the airborne chemical ‘landscape’ shaped by the emissions and wind dispersal. Effective operation of such complex systems needs integrative tools comprising dynamic component modelling. We describe a linked modelling system which reduces uncertainty in monitoring for pest management. The spatial information from surface-layer landscapes and winds is integrated over multiple length and time scales. Larger scale meteorological flows ( ∼ 5 km) disperse pheromones from sources embedded in the surface layer canopy. Smaller scale plume properties ( ∼ 100 m), are downscaled even further to internal plume-fluctuation properties within the plumes ( ∼ 1 cm). Finally, the distributed plume properties are coupled with insect tracking models for pest distribution prediction. The model hierarchy is simple, with the key being the parameters that are passed from one model level to the next. Meteorological models, like the commercially available TAPM system, can model the weather, i.e. the broad scale mean flow over the terrain and determine the local stability over each surface type in the landscape. The complex chemical landscape within the wind fields in the surface layer can then be predicted with Lagrangian particle models of the dispersion of pheromone plumes from canopy releases. Such sources can be from either attracting females or synthetic disruption systems like traps. The internal structure within the plumes, generated by surface layer turbulence, is also explicitly linked with surface layer characteristics of the fine scale turbulence but at even smaller scales because of the tiny dimensions of the source. We adapt classic Monin–Obukhov similarity theory [Stull, R.B., 1988. An Introduction to Boundary Layer Meteorology. Kluwer Academic Publishers, Dordrecht, Netherlands] for modelling plume fluctuations and develop instantaneous random concentration field properties based on inertial range scaling of turbulent processes. The final component is an agent-based insect tracking model that couples directly to the local instantaneous chemical landscape. The intelligent agent makes navigation decisions based on chemical signals to home onto the target source. The coupled system provides an insect mating prediction that accounts for trap catch counts and the influence of weather and cropping systems on these processes. Studies with this model will ultimately allow for a more accurate inference from trap monitoring and for the better management of pests.