Elevated CO sub(2) stimulates N sub(2)O emissions in permanent grassland

To evaluate climate forcing under increasing atmospheric CO sub(2) concentrations, feedback effects on greenhouse gases such as nitrous oxide (N sub(2)O) with a high global warming potential should be taken into account. This requires long-term N sub(2)O flux measurements because responses to elevated CO sub(2) may vary throughout annual courses. Here, we present an almost 9 year long continuous N sub(2)O flux data set from a free air carbon dioxide enrichment (FACE) study on an old, N-limited temperate grassland. Prior to the FACE start, N sub(2)O emissions were not different between plots that were later under ambient (A) and elevated (E) CO sub(2) treatments, respectively. However, over the entire experimental period (May 1998-December 2006), N sub(2)O emissions more than doubled under elevated CO sub(2) (0.90 vs. 2.07 kg N sub(2)O-N ha super(-1) y super(-1) under A and E, respectively). The strongest stimulation occurred during vegetative growth periods in the summer when soil mineral N concentrations were low. This was surprising because based on literature we had expected the highest stimulation of N sub(2)O emissions due to elevated CO sub(2) when mineral N concentrations were above background values (e.g. shortly after N application in spring). N sub(2)O emissions under elevated CO sub(2) were moderately stimulated during late autumn-winter, including freeze-thaw cycles which occurred in the 8th winter of the experiment. Averaged over the entire experiment, the additional N sub(2)O emissions caused by elevated CO sub(2) equaled 4738 kg CO sub(2)- equivalents ha super(-1), corresponding to more than half a ton (546 kg) of CO sub(2) ha super(-1) which has to be sequestered annually to balance the CO sub(2)- induced N sub(2)O emissions. Without a concomitant increase in C sequestration under rising atmospheric CO sub(2) concentrations, temperate grasslands may be converted into greenhouse gas sources by a positive feedback on N sub(2)O emissions. Our results underline the need to include continuous N sub(2)O flux measurements in ecosystem-scale CO sub(2) enrichment experiments.