Coupling soil–plant–atmosphere exchange of ammonia with ecosystem functioning in grasslands

Abstract

The exchange of ammonia (NH 3) between the atmosphere and the land surface is controlled by both atmospheric and land surface processes and can thus be bi-directional. Whether emission or deposition occurs depends on the nitrogen (N) status of the ecosystem. Resistance models have been developed to represent this bi-directional pattern of NH 3 exchange. Major pathways include exchange with leaf cuticles, plant tissues (via stomata) and with the soil surface. However, the parameters that quantify the emission potential of the foliage and ground surface, the NH 3 stomatal compensation point, χ s, and the soil surface NH 3 concentration, χ soil, respectively, are entirely empirical in these models and do not consider the influence of the plant and soil N status. On the other side, grassland ecosystem models simulate the ecosystem N dynamics, but until now NH 3 biosphere–atmosphere exchange was only treated in a very simple manner. A two-layer resistance model for NH 3 exchange has thus been combined with the grassland ecosystem model PaSim in order to link NH 3 exchange with the ecosystem N dynamics. For this purpose, the plant substrate N pool in previous versions of PaSim has been divided between apoplastic and symplastic compartments, and the apoplastic substrate N pool has been linked to the stomatal NH 3 exchange. In addition, soil ammoniacal N (NH x ) has been partitioned between the soil surface and several soil layers, and the soil surface NH 3 exchange has been linked to the soil surface ammonium (NH 4 +). The new combined model has been parameterised and applied to an intensively managed grassland site in Southern Scotland. The comparison with micrometeorological measurements of NH 3 fluxes has shown that the model can qualitatively reproduce the effects of cutting and fertilisation on the net NH 3 exchange above the canopy. In particular, the model reproduces the expected strong coupling of the NH 3 exchange with the dynamics of the apoplastic substrate N pool. However, peak NH 3 emissions are underestimated, and it is postulated that this could be related to the representation of leaf litter emissions from the soil surface in the model, and to the simulated soil NH x dynamics. Nevertheless, this new version of PaSim is a valuable tool for investigating the influence of different management scenarios or of climate change on NH 3 exchange.