The authors responded in a satisfactory manner to the comments of the reviewers. I agree with reviewer 1 that the sampling pattern appears not to cover the entire area. However, within a slightly undulating landscape underlain by glacial till, the variations in feature space will probably be at short distance: along the hillslope and across the heterogeneous soil textures characteristic for glacial tills. The paired samples will be informative for this variability at short distance. A larger scale trend is not to be expected, and therefore a more regular cover of the samples is not strictly required. One has to appreciate the logistic difficulties in taking soil cores to 2 m deep in a large number of fields for which access is not necessarily granted by the farmers. The authors responded to the remarks on the statistical techniques raised by reviewer 2. I would therefore recommend publication of the paper.