Tillage Erosion in Denmark and Its Effects on Soil Organic Carbon Distribution

The EU’s proposed Soil Monitoring Law recognizes soil erosion as major threat to soil health and requests Member States to monitor different soil erosion processes. Amongst the major soil erosion processes tillage erosion has received rather limited attention. Unlike water and wind erosion, whose effects are often easily visible in the landscape, the extent and severity of erosion caused directly by soil tillage only become evident after decades of tillage through spatial variations in soil properties. Tillage redistributes large amounts of soil from convexities to concavities within fields of rolling topography thus driving a spatially heterogeneous evolution of soil nutrient and carbon stocks. Currently, the few tillage erosion models available are based on a modest number of field surveys and tillage tracer experiments, and we are lacking operational tools for monitoring tillage erosion and its impact on soil health in the long-term.

To explore the potential impact of tillage erosion on the crop land in Denmark we have done a scenario analysis comparing the output from a robust soil redistribution model with remote sensing data of topsoil carbon contents. Our study aimed to (1) map tillage-induced soil redistribution across Denmark at a 10-meter resolution and (2) assess its impact on estimating topsoil organic carbon (SOC) content on arable land. Running the WaTEM model with a LiDAR-derived DEM smoothed to different degrees and assuming a typical tillage intensity, we estimated tillage-induced annual soil redistribution rates. We then used Sentinel 2A-derived bare soil composites and other co-variates together with the soil redistribution rates as predictors for mapping SOC via machine learning.

Our modelling results showed that without smoothing the 10-m resolution DEM, 23% of the arable land in Denmark had tillage-induced soil loss rates >2.5 t ha^-1 a^-1 while 12% exceeded 5 t ha^-1 a^-1. At the highly eroding sites, measured plough layer SOC contents obtained from a national survey tended to be lowest. Soil redistribution modelled with the less smoothed DEM showed stronger correlations with the bare soil composite bands in erosional zones. While the bare soil composite was the main predictor for SOC contents, tillage erosion rate was the only other important predictor at a national scale.

In lack of a mechanistic model for mapping the effect of tillage on SOC stock evolution and other soil properties on the arable land across Denmark, our scenario analysis highlights the unsustainability of current intensive tillage practices. To comply with the Soil Monitoring Law the development of operational tools for mapping actual tillage erosion and its impact on soil health must be prioritized.