Incorporating Stochastic Variables in Crop Response Models: Implications for Fertilization Decisions

Most crop-decision models focus on output variation in response to nonstochastic inputs. Little attention has been given to the effects of stochastic factors on the input-use decision. A crop-decision model is developed to examine the impact of soil moisture on the nitrogen fertilization decision. The model is applied to 15 different soil and weather conditions for corn production in Indiana. Based on estimated response functions, the profit maximizing nitrogen application rate varies from 156 to 185 pounds of actual nitrogen (N) per acre. We evaluate the expected economic losses arising from applying to one set of conditions a rate deemed optimal for another.