Abstract
The formation of soils in any region is influenced by many factors such as the parent materials and the secondary materials derived from them, the vegetation and the history of land use. These factors vary from place to place, and they contribute to the spatial variation in properties of the soil. Quantification of the magnitude, location, and causes of spatial variability is an essential, but insufficient ingredient of soil surveys. Soil samples from the 0‐ to 20‐cm depth were taken covering soils in the Asuansi‐Akroso‐Nta‐Ofin compound association (Lixisol, Cambisol, and Fluvisol association) at the study site by following the nested balanced hierarchical sampling technique. This covered distances between 100 and 0.80 m. Standard laboratory analyses were performed to quantify the selected properties, namely, pH, organic carbon, total nitrogen, total phosphorus, exchangeable potassium, and content of sand, silt, and clay. Classical statistics and geostatistical procedures were performed on the data and models fitted to the variability patterns. Physical and the more stable properties, such as sand, silt, and clay, were fitted with spherical variogram models. These models indicate a high level of spatial dependence, therefore, such properties may be said to be fairly stable in the field. On the contrary, chemical properties, such as exchangeable potassium, were fitted with exponential variogram models, indicating that these properties were less stable and showed dependence over longer distances. The scale of variation of the properties ranged between 35 and 62 m. The degree of uncertainty associated with time and space can be reduced by improved documentation of field variability using the tools of geostatistics.
Acknowledgments
The senior author is grateful to the International Centre for Theoretical Physics (ICTP) Trieste, Italy, and the Swedish International Development Agency (SIDA) for financial assistance and hospitality provided while he was on a research visit to the Centre.