Map-based variable-rate manure application in wheat using a data fusion approach

Abstract

Poor manure management practices lead to excess nitrogen (N) and phosphorous (P) in soils that are still common in many countries in Europe. One way to optimize the economic returns and at the same time minimize the environmental impacts of poor manure management is by Variable Rate Manure (VRM) application. This work aimed at the evaluation of potential economic benefits of VRM application, compared to Uniform Rate Manure (URM) application, using advanced sensing, modeling and control technologies. Soil fertility attributes were measured using an on-line visible and near-infrared (vis-NIR) spectroscopy sensor, and crop normalized difference vegetation index (NDVI) was retrieved from Sentinel-2 images in a field of 8 ha with wheat in Flanders, Belgium. The predicted soil properties and the crop NDVI were clustered into a four-class fertility map using the k-mean clustering, each of which has different fertility level. The fertility map was overlaid by parallel strips representing URM, VRM1 and VRM2 treatments. VRM1 intended to apply 40 % more manure than URM to high fertility zones, 20 % more to medium-high fertility zones, 20 % less to medium-low fertility zones, and 40 % less to low fertility zones, whereas VRM2 adopted the opposite approach to that of VRM1. Results showed that VRM2 had increased crop yield by 1.5 % (0.19 t/ha), but reduced applied N by 4.1 % (11.01 kg/ha) and P by 7.1 % (3.72 kg/ha), compared to URM. The VRM1 approach consumed extra N of 5.21 kg/ha (1.8 %) and extra P of 1.63 kg/ha (3.1 %), to achieve a yield increase of 0.31 t/ha with a profit of 40.06 EUR/ha (2.1 %), compared to URM. However, a slightly smaller profit of VRM2 of 36.32 EUR/ha was calculated. Although VRM1 was more profitable than VRM2 by 0.2 % (3.74 EUR/ha), the latter is recommended as the best approach, as VRM1 increased environmental risks in the sense of increasing both N and P applied, by 5.9 % (17.12 kg/ha) and 10.4 % (5.35 kg/ha), respectively.