Total deposition of nitrogen in Swiss forests: Comparison of assessment methods and evaluation of changes over two decades

Atmospheric nitrogen (N) deposition in terrestrial ecosystems is difficult to quantify, especially in forests. In this study, we compared three approaches for determining the wet and dry deposition of nitrogen (total deposition) at 17 intensively monitored forest sites in Switzerland. Specifically, we considered approaches based on: 1) measurements of bulk deposition and throughfall in 2014 (throughfall method); 2) measurements of bulk deposition and measurements of air concentrations of ammonia (NH3) and nitrogen dioxide (NO2), to which deposition velocities were applied, also in 2014 (inferential method); and 3) a model developed for Switzerland at a high spatial resolution, run for the five-year period 2013-2017 (emission based model). In addition, changes over two decades were assessed using continuous measurements of throughfall and bulk deposition. Further, air concentrations of NH3 and NO2 measured in 2014 were compared with concentrations measured in 2000 at 10 of the sites. The three approaches generally yielded comparable estimates of total deposition, with some notable differences at some sites. For both the model and the inferential method, uncertainties were related to the deposition velocities that were applied to air concentrations of N compounds, especially for NH3. The throughfall method provided a minimum estimate of the total N deposition in the forest, but the fraction of the deposited N that is directly taken up by the canopy remained difficult to quantify. Nitrogen deposition has decreased since the mid-1990s at the majority of the sites. However, deposition of the reduced forms of N seems to have stagnated at the local level. Furthermore, N deposition is still too high in comparison with the range of empirical critical loads of N (CLN). The minimum deposition estimated from throughfall exceeds the lower limit of CLN at all sites except those in the Central Alps. Deposition estimated with the model and the inferential method exceeds the lower limit of CLN at all sites and even exceeds the upper limit at several locations.