Author:

Abstract:

Nitrogen fertilisation of arable land is strictly regulated in Belgium and other European countries by the nitrate directive. Farmers are challenged to adjust fertilisation rates in order to obtain good crop yields with a minimal release of nitrogen to the environment. Recommended nitrogen fertilisation rates are a.o. estimated from mineral N stocks in the soil and the expected N mineralisation rates from soil organic matter, crop residues from the previous growing season, fertilising crops and planned organic fertilisations. These mineralisation rates are often average values valid for average weather conditions. However, N mineralisation rates and losses differ within and between the years since temperature and precipitation are important influencing factors. Therefore, there is a need for more accurate methods to estimate the nitrogen dynamics in the soil during the year. This will make it possible to balance nitrogen gift with crop needs and environmental objectives. Several models, predicting nitrogen dynamics in the soil, have been developed. Most of these models have the major disadvantage that they need detailed data which are often not available to the farmer. Therefore, the Louis Bolk Institute and Wageningen University, both located in the Netherlands, developed a practice oriented model called NDICEA. The most important advantages of NDICEA are that the required input parameters are easily available for the farmers and that it uses actual weather data. Several validation exercises in the Netherlands showed that this model predicts the mineral nitrogen dynamics on parcel level quite accurately during the year. The main objective of this study was to validate NDICEA for Belgian conditions. To investigate this some adjustments were made to the soil parameters and eight agricultural fields with different soil textures were monitored between August 2008 and March 2009. Comparing predictions of the adapted NDICEA model to mineral nitrogen measurements in the field revealed 84% agreement for the 0-30 cm soil layer and 73 % for the 30-90 cm soil layer, accepting an error of 20 kg N/ha. The use and implementation of NDICEA in the Flemish agri- and horticulture is recommended provided that further research is able to fine-tune this model.