Using Models and Soil/Plant Observations to Generate Fertilizer Recommendations

Samenvatting

In countries across the world, farmers have access to a fertilizer recommendation manual which details how to take soil and plant samples, and convert the analysis results into fertilization requirements of arable crops. Such a system also exists for potatoes in The Netherlands. However, the state of the art in fertilizing potatoes goes beyond traditional fertilization systems. It makes use of in-season measurements of crop reflectance (with a satellite, drone, or hand-held sensor) to determine crop nitrogen uptake. This is then converted into a nitrogen fertilizer recommendation through the application of a simple crop growth model; this method has been available to farmers for about 10 years. In a new development, current research aims to take crop and soil nitrogen dynamics into account in detail by using a more complex crop growth model, while repeated observations of the crop and advanced statistical methods are used to keep the model synchronized with the farm field. Preliminary results of this method will be presented. The new fertilization method requires access to real-time data, which can only be realized through a dedicated ICT infrastructure to collect, store, and transmit data. However, the most important success factor for the new fertilization system will be the commitment and active involvement of stakeholders: to provide data, and to articulate what kind of recommendations they need in return.