Nitrous oxide emission from Swedish forest soils in relation to liming and simulated increased N-deposition

Fluxes of N2O were studied in a Norway spruce forest in the southwest of Sweden. Three differently treated catchments were compared: Limed (6 t dolomite ha(-1)), Nitrex (additional N-deposition corresponding to 35 kg ha(-1) year(-1), in small doses) and Control (used as control site). The N-retention was still high (95%) after 2 years of N-addition at the Nitrex site when the flux measurements were performed. Each catchment contained both well-drained and poorly drained soils (covered with Sphagnum sp.). The emissions of N2O were in general low with both a high spatial and temporal variation for all three sites. The measured emissions were 25, 71 and 96 (g N2O-N ha(-1) year(-1)) for the well-drained Limed, Control and Nitrex sires, respectively. The average emissions of N2O from the wet areas were significantly higher than the well-drained areas within the catchments. For the wet areas the measured emissions were larger: 90, 118 and 254 (g N2O-N ha(-1) year(-1)) for the Limed, Control and Nitrex sites, respectively. Comparison between treatments showed the wet Nitrex site to have a significantly higher emission than all other sites. The increased N-deposition at the Nitrex site increased the N2O emissions by 0.2% of the added N for the well-drained soils and about 1% for the wet areas, compared with the control site. Since the wet areas represented only a small part of the forest, their larger er emissions did not contribute significantly to the overall emission of the forest. Neither temperature nor water content of the soil was well correlated with the N2O emissions. Soil gas samples showed that most of the N2O was produced below a 0.3-m depth in the soil.