A geostatistical approach to define a soil fertility index based on the main soil macronutrients

Soil fertility is greatly affected by main soil macronutrients such as nitrogen (N), phosphorus (P), and potassium (K). These macronutrients can be used to define a synthetic fertility index to support soil fertilization. The study was aimed to propose a geostatistical approach to define a synthetic fertility index based on factorial cokriging. It consists in quantifying and reducing the spatial variability of multivariate data to only a few factors, related to different spatial scales. Such factors summarize the variability of multivariate data and can be used to divide the field in areas of similar levels of the three macronutrients. Hundred 100 soil samples were collected according to a quite regular grid (20 m x 20 m) from a field of 3.6 ha located in Bilbies district (Egypt). The joint variation of N, P, and K was modelled by a linear model of coregionalization including a nugget effect and two spherical models at short range (42.4 m) and long range (86 m). The joint multivariate variability of N, P, and K in the study area was synthetized by using the first two factor at short and long ranges. The first factor at long range allowed more effectively to delineate the field into different management zones than at short range.