LAUSR

Scientific information on the spatial variability of soil properties is essential for making sustainable soil and environmental management decisions. The spatial variation of soil electrical conductivity (EC), pH, carbonates, gypsum, gravel, clay, silt and sand in the Gavkhooni sub-basin, Central Iran was studied employing conventional statistics, geostatistics and a geographical information system (ArcGIS). By the method of random sampling within blocks, the study area was divided into 100 blocks of 6×6 km2 and in each block surface, samples (0–10 cm) were taken randomly. The EC content showed strong variation (CV>100%), pH and carbonates exhibited weak variation (CV<10%) and the other properties showed moderate variation. Soil parameters such as gravel, silt and gypsum were best fitted by a Gaussian model, sand and carbonates fitted best with a Spherical model and the remaining soil parameters were fitted to the Exponential model. Based on the models, the spatial correlation (range) of the soil properties greatly changed from 5.6 (sand) to 74.5 km (EC). Soil pH, EC, carbonates and clay had strong spatial dependence whereas the other properties had moderate spatial dependence. The spatial distribution maps were prepared using ordinary kriging interpolation. Comparison of the results using statistical methods indicated that kriging technique had satisfactory accuracy in characterizing the spatial variability.