How well does zone sampling based on soil electrical conductivity maps represent soil variability?

Abstract

In zone soil sampling a field is divided into homogenous areas using an easy to measure ancillary attribute (e.g., apparent soil electrical conductivity [ECa]) and a few samples are taken from each zone to estimate the soil characteristics in each zone. This study determined if ECa-directed zone sampling in two fields in northeastern Colorado could correctly predict soil texture and soil organic matter (SOM) patterns of samples taken by a more intensive grid sample method. Each field, which were predominantly Bijou loamy sand (coarse-loamy, mixed, superactive, mesic Ustic Haplargids), and Valentine sand (mixed, mesic Typic Ustipsamments), was divided into three ECa zones and soil texture and OM content in the top 30 cm of soil were measured. There was a significant difference in the soil texture and SOM in both fields between ECa Zone 1 and Zone 3. Logistic regression showed that in both fields, approximately 80% of the grid sample sites in ECa Zone 1 were correctly predicted. Only 50% of the grid sample sites in ECa Zone 3 were correctly predicted as Zone 3 in one field whereas 77% of the grid sites in ECa Zone 3 were correctly predicted in the other field. However, approximately 80% of the samples in the grid sites :10 m from the zone boundaries were classified correctly as compared to the samples that were <10 m from the boundary in which only 50 to 54% were classified correctly. These results support the utilization of ECa-directed zone sampling as an alternative to grid soil sampling if the transition zones arc avoided.