SAMPLING STRATEGIES FOR FERTILITY ON A STOY SILT LOAM SOIL

Spatial variability of Bray P-1 extractable P, exchangeable K, soil pH, organic matter content, bulk density, and penetrometer resistance was studied using geostatistical concepts in a Stoy silt loam soil (fine-silty, mixed; mesic, Aquic Hapludalfs). Directional semivariograms of all surface soil (0-20 cm) properties revealed strong spatial dependence aligned in the direction of farming operation, except bulk density. Longterm cropping and soil management practices oriented towards the direction of farming operation tended to change the spatial structure of soil properties drastically and an unique pattern of spatial dependence was produced in that direction. The results suggest that sampling locations for surface soil properties affected by farming operations be carefully selected. More soil samples should be collected in the direction perpendicular to the direction of farming operations in order to obtain a more reliable representation of the area sampled. Bulk density (0-20 cm) was found to be less affected by farming activities. Due to the relatively undisturbed conditions of the subsurface soil (20-40 cm), no striking pattern of spatial dependence was observed for most soil properties. The results also showed that a reduction in sample numbers was possible with kriging compared to the classical method for achieving the same level of precision.