Development of a Simplified Model for Nitrogen Fertilizer Recommendation for Maize, Wheat, and Sunflower in Northern Greece

Abstract

In many cases, fertilizer crop recommendations in northern Greece are made based on personal opinions rather than scientific background. This may lead to overfertilization and waste of resources. For sound fertilizer recommendations, a model is necessary that would suggest the exact rates of nitrogen (N) based on soil properties of a large number of soil samples. For this reason we developed a model (named N-WHELM), which is a simplified form of a model by Koukoulakis. The output data of our model (the N recommendations) were estimated using data obtained from 50 soil samples, which represented a wide range of soil properties, and were obtained from northern Greece. We tested our model for maize, wheat, and sunflower and evaluated it against the results obtained by the Koukoulakis model. The results showed that the fertilizer recommendations of the N-WHELM model for all three crops were not significantly different from that of the Koukoulakis model. The good relationship between the two models was confirmed by the significant correlations of the two models for all tested crops (R-2 values at the P < 0.001 level). Although the fertilizer recommendations of N-WHELM were not significantly correlated with nitrate (NO3)-nitrogen (N) or other soil properties (when correlated one by one), they did show a strong relationship with soil organic matter as well as a significant relationship when multiple regressions were performed, with all estimated soil properties taken as independent variables. We concluded that the simplified model we developed was successful in recommending N fertilizer requirements similar to those of Koukoulakis.