When dispersal kernels are uncertain, spatial data can be used to infer remote sources of pest and vector population in agroecosystems

Landscape composition and climatic variation can be analyzed to infer processes influential to pathogen and vector abundance at a point of interest, such as crop fields, and to in turn estimate risk to improve management. Substantial research in plant disease epidemiology has addressed the importance of the dispersal kernel to spatial epidemic dynamics, as well as weather variation’s importance to phenological dynamics of several major insect pests. Phenological models of pest abundance are well developed for a multitude of pests; however, similarly developed spatial models must be improved for prediction of cumulative influence on points of interest. Using simulation and nonlinear regression, this study focuses on estimating the relative importance of different land use categories to pest abundance at a point of interest. We relate both temporally resolved land use and weather data to similarly resolved abundance data for Candidatus Liberibacter asiaticus-positive Asian citrus psyllids (Diaphorina citri) from Southern Texas to test the functionality of our analysis in field data. Previous unpublished analysis identifies wind speed as a variable that contributes to proportion of bacterialiferous psyllids in a population, indicating that a spatial process may be contributing to the spread of the psyllid between groves. The relative specificity of Asian citrus psyllids to citrus trees implies that proximity and patch size of citrus groves should be useful characteristics of the landscape for forecasting psyllid abundance, and when combined with quantitative spatial models and weather data, constitute a basis for predicting spatiotemporal variation in risk for both the pest and pathogen.