Open Access

Hans-Werner Olfs, Charlotte Höpker, Herbert Pralle

• To derive fertiliser recommendations for agricultural crops, soil samples should be taken regularly and analysed for important soil characteristics in specialized laboratories. Recently, devices for so-called "in-situ" soil analysis have been used for this purpose. Stenon offers the FarmLab for this purpose, which is equipped with sensors for recording impedance and absorption spectra (NIR to UV spectral range) of soils, the GPS position, some weather data as well as soil temperature and soil moisture. All measurement data sets are stored in a cloud, analysed using appropriate algorithms and then made available via a web application. According to Stenon, this "laboratory-independent soil analysis in real time" eliminates the need to do soil sampling for lab analysis. The FarmLab was used as part of a survey study in western Lower Saxony on 71 farmers fields and on 3 P/K long-term fertilisation experiments. In each case, measurements were carried out in accordance with the FarmLab specifications and soil samples were taken from the topsoil at each of the measuring points for laboratory analysis. After air-drying and sieving to 2 mm, the soil samples were analysed in the laboratory for CAL-P and -K, CaCl2-Mg and the pH value. Due to missing values in the FarmLab measurement, finally only 64 fields could be analysed in this survey. The agreement between the FarmLab and laboratory data sets can all be described as unsatisfactory for the 4 parameters analysed (r2 values: P=0.28, K=0.06, Mg=0.02 and pH=0.04). No reliable relationships between the FarmLab data and the corresponding laboratory data could be determined for either CAL-P or CAL-K on the long-term test plots. The coefficients of determination were very low (all R2 values < 0.2) both for the individual trial and for the total data pool of the 3 trials. While the differentiation of the long-term different P and K fertilizer supply was well represented in almost all cases using the laboratory data, the FarmLab data generally showed almost no differentiation for the 4 fertilisation treatments at the respective location. Obviously, further evaluations are necessary before more concrete statements can be made about the practical suitability of the Stenon FarmLab.