The majority of experiments on soil management and soil fertility uses experimental and statistical procedures which require statistically independent samples and uniform experimental plots. This hypothesis can only be satisfied and verified in practice with geostatistical calculations such as semivariograms. The objective of this paper was to study the spatial variability, within an experimental plot, and to show the use of geostatistical analysis. The field selected for this study is located at the Centro Experimental de Campinas, Instituto Agronômico, Campinas (SP), Brazil, on a Dusky Red Latosol, which had been on conventional tillage for the previous 10 years. A 30 m by 30 m plot was marked on a square grid of 5 m, resulting in 49 sampling points, in which bulk soil samples were collected at the depths of 0-25 cm and 25-50 cm, taken to the laboratory, air dried, sieved on the 2 mm sieve, and submitted to routine chemical and particle size analysis in order to obtain the clay content, silt content, delta pH, sum of bases, cation exchange capacity, and base saturation. Geostatistics was used for the spatial variability analysis. The coefficients of variation found indicate higher variability for all properties at 0-25 cm than at 25-50 cm, probably due to soil tillage depth. Spatial dependence was also stronger at 0-25 cm than at 25-50 cm depth. It was concluded that the spatial variability for the chemical properties was high considering the plot size of 30 m by 30 m and that random sampling would fail to detect it.
soil; sampling technique; geostatistics; clay; silt; pH; bases content; cation exchange capacity; base saturation
