SPATIAL DEPENDENCE OF SOIL POTENTIALLY AVAILABLE NITROGEN IN TWO CONTRASTING SITES

Soil nitrogen (N) variability affects the effective application of N fertilizers. The objective of this research was to determine and assess the spatial dependence of soil N in two different soils using spatial sampling and geostatistics fundamentals. An optimized nested sampling design was applied in two sites, using 16 main stations from where soil samples (0 - 23 cm) were obtained at 1.5, 4.5, 13.5, 40.5, and 121.5 m, at random directions. Soil samples were collected from an Andisol and an Inceptisol (2 x n = 98), and the potentially available N (PAN) was measured by anaerobic soil incubation for 7 days at 40oC (PAN = NH4+ day 7 - NH4+ day 1). The PAN variance components from each distance were calculated by residual maximum likelihood (REML), and then accumulated to plot an approach to the variogram. The results showed a wide range of values for PAN at both sites. The Andisol site registered high PAN levels, with a mean value of 32.92 mg NH4+ kg soil(-1), which was five times larger than in the Inceptisol soil, probably associated with a larger soil organic matter content. The limit of the spatial dependence was 13.5 m for the Andisol, and 40.5 m for the Inceptisol soil, corresponding to the 59% and 35% of the process variance, respectively.