Search

Establishing a soil sampling plan is one of the most important stages for providing the detailed information required for sustainable land management. This study aims to investigate the most appropriate sampling density and proportion for the characterization of Oxisol attributes on different materials, using the magnetic susceptibility (MS). The studied area is located in São Paulo State, Brazil, in Guatapará Town and its size is 380 ha. To evaluate the subsampling effect in the variability description, 60 samples were collected from part of the area (150 ha) in the density of one point per 2.5 ha, using both simple sampling and subsamples with three points. Different sampling densities were obtained (3; 4; 5; 6; 7; 8; and one point per 9 ha) by the exclusion of points. The evaluated attributes were: clay content; base saturation and MS at 0.00-0.25 and 0.25-0.50 m depths. There were no differences among spatial variability maps generated from either simple or subsamples. This information might precede a change in the sampling system for the agricultural industry. The minimal sampling density required so that there is no loss in the spatial variability representativeness was one point per 7 ha. MS may be used to upgrade clay content estimates in the sampling proportion of three points of MS for per clay point (one point per 2.5 ha for MS: one point per 7 ha). [ABSTRACT FROM AUTHOR]

Spatial sampling designs used to characterize the spatial variability of soil attributes are crucial for science studies. Sample planning for the interpolation of a regionalized variable may use several criteria, which could be best selected from an estimated semivariogram from a previously established grid. The objective of this study was to optimize the procedure for scaled semivariogram use to plan soil sampling in sugarcane fields in the Alfisol and Oxisol regions of Jaboticabal Town in São Paulo State, Brazil. A scaled semivariogram for several soil chemical attributes was estimated from the data obtained from two grids positioned on a sugarcane field area, sampled at a depth of 0.0-0.5 m. The research showed that regular grids with uniform intervals did not express the real spatial variability of the soil attributes of Oxisols and Alfisols in the study area. The calculated final sampling density based on the scaled parameters of the semivariogram was one sample for each 2 ha in Area 1 (convex landscape) and one sample for each 1 ha in Area 2 (linear landscape), as indicated by SANOS 0.1 software. The combined use of the simulation programs and scaled semivariograms can be used to define sampling points. These results may help in soil fertility mapping and thereby improve nutrient management in sugarcane crops. [ABSTRACT FROM AUTHOR]