Abstract: To explore the reason for reported high field fluxes of nitrous oxide (N2O) from temperate soils in winter and early spring, we investigated the temperature response of denitrifier N2O production and reduction in soil from three arable field sites along a temperature transect reaching from Finland over Sweden to Germany. Process rates were determined in anaerobic slurries with or without added NO3-, N2O and C2H2 at 0, 5, 10, 15, and 20C (and 30C in one experiment). The experiments were conducted immediately after the soils had become anaerobic, and after a long (48 h) anaerobic pre-incubation with excess of carbon and electron acceptors. All denitrifying enzymes were found to be active in the soil at onset of anaerobiosis. Significant levels of N2O production and reduction occurred at 0 8C, both at onset of anaerobiosis and after the 2 days anaerobic pre-incubation. Temperature response of N2O production and reduction could be fitted to an Arrhenius function in the range 5-20 °C, yielding apparent activation energies between 28 and 76 kJ mol -1. The estimated activation energy of the N2O reduction was found to be similar or lower than that for N2O production. High field N2O fluxes in winter and early spring could thus not be explained by the temperature sensitivity of the two processes. However, major deviations from the regular Arrhenius response were found for two soils at near freezing temperature. The rates measured at 0 °C were much lower than those predicted by the Arrhenius function based on data in the temperature range 5-20 °C. Low temperature may thus exert a particular challenge to denitrifying communities for some reason, and the effect was found to be most severe for the N2O reduction process. When such a breakdown affects N2O reductase to a greater extent than the N2O producing enzymes (NO3-, NO2-, and NO reductase), as was found in our soils, it will result in high N2O fluxes at low temperature. The temperature response of the estimated net N2O emission potential (based on measured N2O production and reduction rates) differed significantly between the three sites, indicating inherent differences between their microbial communities.
