A geostatistical method for soil salinity sample site spacing

Abstract A calibrated Four-Electrode Probe (FEP) was used for inexpensive and indirect determinations of salinity-sensor Electrical Conductivity (EC) in a plot at Cauto Valley, Cuba. Two transects were made in the North–South (N–S) and East–West (E–W) directions. Laboratory measurements of soil EC were also made from samples taken on a 50-m spaced square grid. A linear semivariogram was obtained for the salinity-sensor EC measurements at the E–W transect, which agrees with the topographical slope of the plot and with the expected soil salinity variation. It also coincides with the spatial structure of laboratory-measured soil EC. A cross-validation analysis has shown that EC semivariograms obtained from FEP measurements can characterize the soil EC spatial variation in a similar way as semivariograms of laboratory-measured soil EC. Thus, the distance between samples for soil salinity maps can be based on the semivariogram's range of salinity-sensor EC measurements.