When reducing the negative impact of agriculture on the environment, one major approach is to substantially decrease the nitrate contamination of the groundwater originating from mineral and organic fertilization. However, efficient plant production relies on sufficient nutrient supply. Finding the balance between appropriate plant nutrition and ground water conservation is a challenge. Appropriate nitrogen fertilization relies on the measurement of mineralized nitrogen (Nmin) - the total amount of nitrate and ammonium - in the soil. Taking soil samples with an auger and subsequent chemical analyses in a laboratory is state of the art. However, this process is time-consuming and often time-delayed. The Stenon FarmLab, a sensor spade, is a new device which claims to measure soil Nmin and other soil parameters in real-time with spectral methods. The aim of this study was to validate the Nmin measurements of the Stenon FarmLab with the common laboratory method. The authors conducted a series of 20 measurements consisting of a varying number of individual values per measurement. In total, 211 individual values on 15 different field sites in three regions in Bavaria, Germany were analyzed. Reference samples for laboratory analysis were taken simultaneously with a sampling auger. Only 181 of the 211 individual values were considered in the evaluation as Nmin contents of less than 42 kg N ha−1 and greater than 189 kg N ha−1 exceeded FarmLab's measurement range. The results showed that the Stenon FarmLab overestimated Nmin in 75 % of the cases in comparison to the laboratory analysis. On average, the mean values of the sampled sites differed by 38 kg N ha−1 showing a 69 % mean deviation from the laboratory. The overall coefficient of determination (R2) was 0.3 for individual measurements and 0.66 for the mean values of a site. In conclusion, the Stenon FarmLab is a good approach for a more time-saving method regarding relative differences within or between fields. However, absolute values measured with the Stenon FarmLab, which are required for demand-driven fertilization, are not accurate enough: the system needs to be further improved to match reference methods.
Read full abstract