Usefulness of Models in Precision Nutrient Management.

Modern agriculture increasingly applies new methods and technologies to increase production and nutrient use efficiencies and at the same time reduce leaching of nutrients and greenhouse gas emissions. GPS based ECa-measurement equipment, ER or EM instrumentations, are used to spatially characterize soil condition by mapping a variety of physical-chemical properties including salinity, water content, texture, bulk density. vis–NIR spectroscopy shows the potential for use directly for the characterisation of soil quality, or soil fertility as the spectra contain information on soil organic and mineral composition. Mapping of crop status and the spatial-temporal variability within fields with red-infrared reflection are used to support decision on split fertilisation and more precise dosing. The interpretation and use of these various data in precise nutrient management is not straightforward, especially as the various soil and crop state variables or indicators are a result of several processes integrated over time. The present paper explore the possibility to use a physically based model to interpret the spatial variable data and if the model sensitivity to changes in input my lead to a valid and useful output, such as spatial variable need for nitrogen fertilisation. Sensitivity analysis showed that the model was sensitive to the soil texture variation often found within fields, but especially increases in soil organic matter content and the depth of the A-horizon resulted in significant upsurge of mineralisation. However, whether the crop would benefit from this depended to a large extent on soil hydraulic conductivity within the range of natural variation when testing the model. In addition the initialisation of the distribution of soil total carbon and nitrogen into conceptual model compartments was critical and difficult. Further several years of historic information on crops and crop management for the given field were needed to take into account nitrogen carry over effects.