In a regular grid with 31 points located with GPS, samples were taken and based on geostatistics analysis the spatial variability of soil properties and their relationships with yield of a mango crop were studied. It was found that soil properties and crop yield presented a high, within-field, spatial variability. The range of the semi variance models varied from 10 to 192 meters, with the exception of pH and Ca, the adjusted models showed nugget/sill fraction lower than 30%, thus indicating that most of the process is explained by the structural variation and then better interpolation results could be expected. K, Ca, CIC, Ca+Mg/K fraction, slope, and altitude were the most important factors affecting crop yield. The studied method in this research is useful to define management zones within fields and some used criteria like altitude and slope are important for the zoning of larger areas and can be computed from a digital elevation model.
spatial variability; geostatistics; mango crop
