This study aimed to analyze the spatial variability of the chemical attributes and texture of a red latosol. The study was carried out on a property in the town of Sidrolândia, Matto Grosso do Sul, which produces soybeans, corn and cotton, using crop rotation and a no-tillage system. In an experimental area of approximately 90 hectares, samples were taken of the soil at 181 geo-referenced points. The samples were each composed of four single samples, representing depths of from 0 to 0.2 m. The chemical and texture of the soil were measured. Initially a discrepancy analysis was performed followed by descriptive and geostatistical analyses, to characterize the samples and identify the spatial dependence of the attributes studied. No spatial dependence was detected for the attributes: active water acidity, aluminum, potential acidity, base saturation and organic matter. The best estimates of values for non-sampled locations were observed for the physical properties of the soil, which presented parameters which best fit the variograms and cross-validation. The geostatistical techniques used allowed adjustment of the theoretical models that best represented the experimental semivariance, thus enabling the construction of thematic maps of spatial distribution for the values of the attributes of the studied soil.
Oxisol; Precision agriculture; No-tillage
