Grades amostrais para análise da variabilidade espacial de atributos do solo e a produtividade da soja

Abstract

Precision agriculture comprises tools and technologies for the rational use of agricultural inputs. Georeferenced soil sampling techniques are employed to evaluate the spatial variability of soil chemical attributes and particle size, which are related to crop productivity. The aim of this study was to assess regular sample grids for determining the spatial variability of soil attributes and their correlations with NDVI (Normalized Difference Vegetation Index) and soybean productivity. The experiment was conducted in a 224 ha plot of a commercial farm located in Maracaju – MS, with Dystroferric Red Latosol. Sample arrangements were generated using a regular grid with one sample per hectare (1:1), with reduced density in subsequent evaluations (1:2, 1:3, 1:4, 1:5, 1:7, 1:10). Calcium (Ca), Magnesium (Mg), pH CaCl2, H + Al, Organic Matter, Sand, Silt, and Clay contents were evaluated, and the results underwent descriptive statistics, geostatistics, ordinary kriging, and kappa index. Additionally, the NDVI of the soybean crop for 2022/2023 was assessed based on available Sentinel 2 satellite images and correlated with soil attributes. According to the results, the soil in the evaluated area is clayey, and all sample grids showed satisfactory correlation with soybean productivity, which varied with the Clay content in the area. The 1:3 sample grid was satisfactory for assessing pH CaCl2 and Calcium levels. The 1:4 and 1:7 grids were efficient for evaluating Organic Matter and Magnesium, respectively. For aluminum, all grids could be used, except the 1:3 grid.