Soil electrical conductivity map quality

Soil electrical conductivity (EC) maps must be of adequate quality to be of value for site-specific management. The objective of this paper was to assess the quality of EC maps created with different sampling intensities and interpolation procedures. This study was conducted for fields with limestone soils overlain by loess from the Inner and Outer Bluegrass physiographic regions of Kentucky. The two Inner Bluegrass fields were in tall fescue (Festuca arundinacea Schreb.), and the Outer Bluegrass field was in a no-till corn (Zea mays L.)-soybean (Glycine max (L.) Merr.)-wheat (Triticum aestivum L.) rotation. Bulk soil EC was measured by direct contact along transects when soils were dry. Prediction and validation datasets were collected in perpendicular directions. The data were interpolated with ordinary kriging and inverse distance weighted (IDW) interpolation. Map quality was quantified with prediction efficiency, a reduction-in-error indicator. Log-normal and Box-Cox transformations did not improve spatial predictions substantially. Across locations, sampling intensities, and EC measurements, IDW with a distance exponent of 1.7 (IDW_1.7) maximized prediction quality, and prediction efficiency values were 4.0% greater for IDW_1.7 than ordinary kriging interpolation. For deep and shallow EC at each location, models were developed that described the relationship between sampling intensity and prediction efficiency (minimum r² = 0.82). These models were used to demonstrate that by collecting more intensive measurements in fields with greater EC variability, as recommended by the EC sensor manufacturer, the quality of EC maps could be substantially improved. Spatial variability should be assessed during a preliminary survey of EC variability to aid in determining the appropriate sampling interval.