The objective of this work was to analyze the spatial variability of fertility attributes in an Oxisol under tillage to cotton cultivation in Cristalina, GO, aiming the adoption of precision agriculture in the area management. The soil was collected using a sample grid, which consisted of 90 points (80 x 80 m), distributed into 57.6 ha. Were determined contents of P, K, Ca, Mg, S, Na, Cu, Mn, B, Zn, Fe, organic matter (OM), pH, sum of bases (SB), base saturation (V), aluminum saturation (m), as well as the apparent electrical conductivity of soil (EC) measured at 30 and 90 cm depth. Classic descriptive analysis was conducted and then, semivariograms were modeled for all attributes and krigging maps. The spatial dependence of analyzed soil properties indicates that the spatial variability should be considered during the planning of soil sampling and also to soil management practices. Concerning to precision agriculture adoption and the application of inputs in variable rates, it is required to do a careful study of the soil fertility variability, creating maps based on fertility classes for each attribute, allowing to identify for which attributes this management practice should be adopted.
kriging; geostatistics; variable rate application
