Parameterizing Bioeconomic Models of Soybean Pest/Pathogen Management with Small Plot Data.

We have been developing a general structure for a stochastic bionomic model of pest and pathogen management for annual row crops that can be parametrized with data from small-plot experiments that production scientists commonly conduct. The model is sufficiently general to apply to a variety of pests, pathogens and crops, with applications here focusing on management of soybean aphid (Aphis glycines) and white mold (Sclerotinia sclerotiorum) in soybean (Glycine max). The model begins with a probability distribution for pest pressure with parameters estimated using data from the untreated control plots. Pest pressure can be measured using pest population densities, disease severity or crop injury due to a pest, such as cumulative aphid days (CAD) soybean aphid, the disease severity index (DSI) for white mold, or the node injury scale (NIS) for rootworm (Diabrotica spp.) larval feeding damage in corn. The probability distribution of pest pressure with control is estimated as a density function with parameters conditional on the initial pest pressure and that vary among treatments. For example, the mean and standard deviation of CAD with an insecticidal seed treatment would be some proportion of the CAD without treatment. The yield benefit of a control treatment then depends on the reduction in pest pressure achieved by the treatment, with a greater average benefit generated by a greater reduction in pressure. Finally, the economic value of these yield benefits are determined using pest/pathogen-free crop yields, crop price, and treatment cost information. Because the model is stochastic, the break-even probability for a treatment can also be estimated.