Nitrous oxide emissions from agricultural land use in Germany - a synthesis of available annual field data

The nations that have ratified the Kyoto Protocol must set up an appropriate national inventory on N2O emissions from agricultural land use, in order to report properly on the achievements made in reducing greenhouse-gas emissions. The search for the appropriate method is a controversial topic as it is subject to high uncertainty in particular associated to the upscaling from site measurements. In this study, all available data from Germany on annual N2O-emission rates derived from field experiments of at least an entire year are summarized. From each study, only differences in soil properties on N input qualified as an individual data set. Under these premises, 101 treatments from 27 sites were found equally spread across Germany. The annual N application ranged from 0 to 400 kg N ha(-1) and the annual emission rates from 0.04 to 17.1 kg ha(-1). Annual emission factors (EFs), uncorrected for background emission, varied considerably from 0.18% to 15.54% of N applied. There was no nationwide correlation found for the relationship between N2O losses and N application, soil C, soil N, soil texture, or soil pH. However, site-specific trends in the relationship between emission factor and mean soil aeration status, as expressed by the soil type and/or mean climatic conditions, were revealed. Regularly water-logged soils were characterized by low emission factors as were soils from the drier regions (< 600 mm y(-1)), whereas well-aerated soils from the frost-intensive regions showed exceptionally high emission factors. Since purely physical and chemical parameterization failed to describe N2O emissions from agricultural land use on the national scale, there must be a biological adaptation to mean site conditions, i.e., different microbial communities react differently to similar actual conditions in terms of N2O dynamics. Regardless of the point of view, the chapter on N2O Soil dynamics cannot be closed yet, and new additional model concepts, process studies, and field measurements are needed.