Changes in organic matter contents and carbon stocks in Dutch soils, 1998–2018

Abstract

Changes in soil organic matter (SOM) content and soil organic carbon (SOC) stock in the 0–30 cm and 30–100 cm soil layers between 1998 and 2018 in the Netherlands were estimated by repeated sampling of 1152 locations in the Soil Sampling Programme (SSP). These locations were selected following a stratified simple random sampling design. We discuss various barriers we met: restricted accuracy of information on soil bulk density, uncertainties due to positional errors, differences in sampling support, and changes in laboratory analysis methods since 1998. Domains of interest such as mineral soils were defined either on the basis of the stratification of the SSP sample (geomatching) or on the basis of soil profiles observed at the selected locations (classmatching). The mean SOM content changed significantly in the 30–100 cm layer (-17.68 gkg-1) in the entire area of interest (non-built-up area in the Netherlands) between 1998 and 2018 (at a 5% significance level). A decrease in SOM content between 1998 and 2018 could be shown for the 0–30 cm layer in mineral soils under cropland if classmatching was applied (at a 5% significance level), but no change could be shown in this layer in the remaining domains of interest, whether geomatching or classmatching were applied. For the 30–100 cm layer in mineral soils, significant changes in mean SOM content were shown by classmatching: −8.59 gkg-1 under cropland and −4.75 gkg-1 under grassland. The calculations indicate that SOC stocks decreased between 1998 and 2018 in both the 0–30 cm and the 30–100 cm layer of mineral soils under both cropland and grassland. The accuracy of the bulk density data needs to be improved in future measurements to increase the accuracy of calculations of the SOC stock changes.