Text S2. Comparison of data sets for nitrogen input from agricultural sources.

To estimate systematic differences between both national N surplus time series, we selected the European estimation for croplands [kg] by West et al. (2014) as reference. Because this data set covers only the years around 2000 with a 5 arc minute resolution, we could not take it as a consistent substitute for the national data.
A comparison of the national with the European excess N showed a good fit (23 vs. 19 kg N ha-1 yr-1) for the French data (Table S2.), but an underestimation of the national estimations (34 vs. 47 kg N ha-1yr-1) for Germany by 28%. The estimates of retention in Germany may therefore be slightly higher than assumed taking the national N input data. But all in all, the methodological offset between both national data sets can be assumed as acceptable due to the comparison with the reference by West et al. (2014) and under consideration of uncertainties in the national as well as European estimations. Furthermore, a comparison of the national N diffuse input in comparison to the European N excess data promised a more consistent estimation for N input per hectare (Figure S2.).