Precision Support Unit Swedish Farmers ' Cooperative

The performance of different methods for mapping clay, soil organic matter content (SOM) and pH(H2O) within a field, using information on soil electrical conductivity was assessed. Soil samples were taken from 39 locations within an arable field in southwest Sweden. Electric conductivity was measured densely using EM38. The bimodally distributed soil electrical conductivity data were used to divide the field into two zones. Averages within these zones were used for mapping the soil variables. This method was compared with a linear regression model derived from the relationship between soil electrical conductivity and soil data as well as with cokriging with soil electrical conductivity as a covariable and with simple interpolation with inverse squared distance weighting. For pH(H2O) the mapping quality was not improved with any of the above methods, compared to using a simple field average. The map of SOM created through cokriging was of similar quality as the division into two homogenous zones. Both were better than the linear regression model and inverse distance weighting, which gave no better quality than a simple field average. Cokriging worked best for mapping clay content followed by the linear regression model, the division into zones and finally inverse distance weighting