Evaluation of the PNM Model for Simulating Drain Flow Nitrate-N Concentration Under Manure-Fertilized Maize

Abstract

Mathematical models may be used to develop management strategies that optimize the use of nutrients from complex sources such as manure in agriculture. The Precision Nitrogen Management (PNM) model is based on the LEACHN model and a maize N uptake/growth and yield model and focuses on developing more precise N management recommendations. The PNM model was evaluated for simulating drain flow nitrate-nitrogen (NO3-N) in a 3-yr study involving different times of liquid manure application on two soil textural extremes, a clay loam and a loamy sand under maize (Zea mays, L.) production. The model was calibrated for major N transformation rate constants including mineralization, nitrification and denitrification, and its performance was tested using two different calibration scenarios with increasing levels of generalization: (i) separate sets of rate constants for each individual soil type and (ii) a single set of rate constants for both soil types. When calibrated for each manure application treatment for each soil type, the model provided good simulations of monthly and seasonal drain flow NO3-N concentrations. The correlation coefficient (r) and Willmott’s index of agreement (d) ranged from 0.63 to 0.96 and 0.72 to 0.92, respectively. The calibrated model performed reasonably well when rate constant values averaged over manure application treatment for each soil type were used, with r and d values between 0.54 and 0.97, and 0.70 and 0.94, respectively, and greater accuracy for the clay loam soil. When rate constant values were averaged over manure application treatments and soil types, model performance was reasonably accurate for the fall time manure application on the clay loam (r and d of 0.60 and 0.91 and 0.72 and 0.92, respectively) and satisfactory for the spring time on the clay loam and the fall and spring times for the loamy sand soil (r and d between 0.56 and 0.90 and 0.58 and 0.84, respectively). The use of the model for predicting N dynamics under manure-fertilized maize cropping appears promising.