AmaizeN: A decision support system for optimizing nitrogen management of maize

Abstract

AmaizeN is a decision support system to help maize growers schedule nitrogen (N) fertilizer applications for site-specific maize crops. It forecasts crop yields and N-fertilizer application rates for potential yield and best economic returns, and predicts the consequences of user management decisions. It takes into account both crop production and environmental impact. In this article we describe the system functionality and underlying crop models, and the system validity and effectiveness evaluated in 16 field trials covering a wide range of weather and soil conditions. At each trial site crops received either two or four N-fertilizer application rates, including one rate recommended by AmaizeN. The AmaizeN-predicted maize yields matched field measurements well ( r 2 = 0.77; p < 0.001 for silage, and r 2 = 0.55; p < 0.001 for grain), and gave a reasonably good indication of silage crude protein content ( r 2 = 0.28; p < 0.001) and silage harvest date ( r 2 = 0.71; p < 0.0006). The system was also capable of estimating N-leaching during the cropping season and predicting residual soil mineral-N at the end of the season ( r 2 = 0.47; p < 0.001), but more effort is needed to improve the accuracy of some predictions. In all instances the AmaizeN-recommended N-fertilizer strategy was more efficient than the growers’ practice. Recommended N-fertilizer rates were on average 85 kg ha −1 less than conventional application rates across 10 crops, with no yield reduction. Its recommended higher-than-conventional application rate at another crop brought about a significant yield increase. System development was guided by an industry user group who requested the decision support system interface to be ‘simple and easy to use’. To ensure user adoption of the system some compromises in system prediction accuracy were required. Local agricultural production conditions were also incorporated.