DESIGN OF OPTIMIZING THE APPLICATION OF THE MANURE IN HIGH NATURAL VALUE (HNV) GRASSLANDS, IN SUSTAINABLE AGRICULTURE, BASED ON NDVI

Abstract

In many countries, especially in Eastern Europe, there is much discussion about how grasslands are managed in order to prevent pollution of soils, rivers, lakes and to maintain and conserve wild flora and fauna. These objectives favor the maintenance of high biodiversity in High Natural Value (HNV) grasslands through their extensive use, with a low intake of organic fertilizers in line with EU CAP policies. In Romania, in the High Natural Value (HNV) grasslands, subsidies are allocated for "environmentally friendly practices", through the Agri-environment and climate program, but the amount of organic fertilizer is restricted up to 40 kg nitrogen active element/ha regardless of the type fertilizer) but also their administration. Currently, through remarkable advances in areas such as GIS and remote sensing, applications for agriculture provide complex information with high accuracy, but also the ability to provide predictions for various parameters. In this context, the paper aims to optimize the application of manure in HNV grasslands, in accordance with Agri-environment policies, by improving and adapting in practice the OneSoil application based on NDVI map generation from Sentinel satellite images. The validation of the "adjustments" brought to the application was done by field visits with the Phantome 4 UAV equipment. An experimental site was considered an HNV grassland located in Brădişoru de Jos (ATU Oraviţa) which has an area of 391 ha. In the study area the vegetation is mosaic, the grassland being fragmented by hedges, forest curtains and trees in clumps that will be excluded from fertilization. The usable area for fodder at the level of 2020 was 265.41 ha, but in the absence of cleaning works in the coming years, there is a risk of substantial reductions. The proposed optimization model is finally presented in the form of a “coverage” map that indicates both the required quantity, differentiated according to the vegetation characteristics suggested by NDVI, and the spatial location of these quantities on subzones. Thus, on an area of 60.73 ha, 4 t.ha-1 of manure are required, on 94.66 ha it is necessary to administer 5 t.ha-1 of manure, and on an area of 110.02 ha, where the vegetation of the grassland suffers, it is necessary to administer 8 t.ha-1 of cattle manure. The major advantage of this model is that the user can import, on different devices (phone, tablet or laptop) vector files with plot outline or GPS points for location, can differentiate the dose of fertilizer on the surface of the plot, under specific conditions and has at its disposal the map with the "location" of the different quantities of fertilizers.