Dynamics of atmospheric ammonia exchange with intensively-managed grassland
Continuous measurements o f atmospheric ammonia (NH3) exchange were conducted for a period o f 19 months (May 1998-November 1999) over intensively managed grassland (cut twice for silage and grazed) in southern Scotland using the aerodynamic gradient method. The mean NH3 concentration and flux for the whole measurement period were 1.52 pg m ' and 13.9 ng n f s’ , respectively.Enhanced emissions o f NH3 were observed following four separate grass cutting events (June 1998, August 1998, June 1999 and May 2000) with peak emissions o f 380, 200, 539 and 508 ng m’2 s '1, respectively. The magnitude o f these emissions was up to an order o f magnitude greater than the emissions observed from the grassland prior to cutting. Enhanced NH3 emissions from cut grassland have been observed, but not quantified prior to this study.The NH3 exchange was bi-directional with large diurnal and seasonal variation, which was strongly linked to grassland management in addition to meteorological conditions. The grassland varied from being a net sink for NH3 during winter months (-6.0 g NH3-N ha' 1 d '1) and prior to cutting o f the grass (-4.9 g NH3-N ha' 1 d’1) to being a net source after the grass was cut (29.3 g NH3-N ha' 1 d '1) and after nitrogen fertilisation (153.6 g NH3-N ha' 1 d’1). Net emission was also observed during grazing periods (33.0 g NH3-N ha' 1 d '1). The pattern o f NH3 exchange was similar for 1998 and 1999.The net annual budget o f NH3 exchange for the grassland for May 1998-April 1999 was emission o f NH3 o f 1.9 kg N ha' 1 y r'1, equating to 1.6% o f the fertiliser N applied. The gross emission flux for the year was 4.2 kg N ha' 1 yr’1. Scaling up these gross emissions across the whole o f the UK improved grassland (60,500 km2) would lead to 25 kt NH3-N, equivalent to 9.5% o f the UK total emissions. These results indicate that the gross emission from all processes in fertilised grassland, including emissions from fertilisation, grazing and from cutting, make a significant contribution to the NH3 emission budget o f the UK.A two-layer canopy compensation point resistance model was applied to the NH3 measurements. Close agreement between measured and modelled fluxes was obtained by introducing seasonally dependent functions o f the foliar and ground layer emission potentials (rs, rg) for key periods. This is a significant improvement on the current use o f constant emission potentials within national deposition models.In addition to the resistance modelling approach, a fully dynamic grassland ecosystem model (PaSim) was applied to the NH3 measurements. In PaSim, the emission potentials and NH3 exchange are linked functionally to dynamic plant and soil N pools. Scenarios o f changing climate and management were explored with the model. The simulated NH3 emissions did not follow the expected thermodynamic response to a rise in temperature and demonstrated the complexity o f the ecosystem level response o f NH3 exchange to climate change. Simulated NH3 emissions were reduced by 15% by delaying the timing o f fertiliser applications by two weeks, indicating the potential o f this measure as an NH3 abatement option.The first major intercomparison o f NH3 gradient measurements was conducted during a field campaign over intensively managed grassland in Germany. Enhanced emissions o f NH3 after grass cutting were also observed at this site, providing additional corroboration for this emission source, while the temporal pattern of exchange was similar to that observed in Scotland.