Experimental determinations of isotopic fractionation factors associated with N2O production and reduction during denitrification in soils

Abstract

Quantifying denitrification in arable soils is crucial in predicting nitrogen fertiliser losses and N2O emissions. Stable isotopologue analyses of emitted N2O (δ15N, δ18O and SP=15N site preference within the linear N2O molecule) may help to distinguish production pathways and to quantify N2O reduction to N2. However, such interpretations are often ambiguous due to insufficient knowledge on isotopic fractionation mechanisms. Here we present a complex experimental approach to determine the net fractionation factors (η) associated with denitrification. This determination is based on three laboratory experiments differing in their experimental set-up and soil properties. Static and dynamic incubation techniques were compared. All available methods for independent determination of N2O reduction contribution were used, namely, N2-free atmosphere incubation, acetylene inhibition technique and 15N gas-flux method.For N2O production: (i) the determined difference in δ18O between soil water and produced N2O vary from +18‰ to +42‰ and show very strict negative correlation with soil water saturation; (ii) the determined η15N of N2O production vary from −55‰ to −38‰ and the fractionation decreases with decreasing substrate availability; (iii) the determined SP of produced N2O vary from −3‰ to +9‰. For N2O reduction: (i) the determined η18O and η15N of N2O reduction vary in very wide ranges from −18‰ to +4‰ and from −11‰ to +12‰, respectively, and depend largely on the differences in experimental setups; whereas (ii) the determined ηSP of N2O reduction shows a very consistent value with all previous studies and varies in a rather narrow range from −2‰ to −8‰. It can be concluded that η values of N2O production determined during laboratory incubations yield only roughly estimates for respective values expectable under field study conditions. η18O and η15N associated with N2O reduction may vary largely, probably depending on spatial and temporal coincidence of N2O production and reduction, and are hence not yet predictable for natural conditions. However, the ηSP of N2O reduction appeared to be relatively robust and a most probable value of about −5‰ can be used to constrain N2O reduction based on SP of soil emitted N2O.