Summary: | Tests to evaluate competition between plants, in general do not consider the soil spatial variability, nevertheless, the uniformity rarely is present and can not be assumed without verification. The aim of this work was to apply geostatistics to verify the spatial variability of the soil in an experimental field near Ribeirão Preto, SP, with sugarcane variety IACSP93-3046. Ninety seven geo-referenced samples, placed at distances of 10 m, were collected for soil chemical analysis, soil resistance to penetration and biometric evaluation of the sugar cane plant included number of tillers, stalk diameter, stalk height, estimated productivity (TSSe) and root density. Geostatistics has been applied by means of semivariogram, data interpolation via kriging and iso-line maps creation. The results have shown soil spatial dependence for most of the evaluated attributes. The spots, within the area, with low resistance to penetration and low soil density have shown the largest number of tillers and the largest root development of sugar cane. For soil chemical attributes, there was spatial dependence showing higher concentration of nutrients in the central area. The sampling allowed a good representativity of the spatial dependence of soil and plants, making it possible to eliminate the randomness hypothesis for the placement of the plots in this area.
|