Alteration of nitrous oxide emissions from floodplain soils by aggregate size, litter accumulation and plant–soil interactions

Abstract. Abstract. Semi–terrestrial soils such as floodplain soils are considered potential hotspots of nitrous oxide (N 2 O) emissions. Microhabitats in the soil, such as within and outside of aggregates, in the detritusphere, and/or in the rhizosphere, are considered to promote and preserve specific redox conditions. Yet, our understanding of the relative effects of such microhabitats and their interactions on N 2 O production and consumption in soils is still incomplete. Therefore, we assessed the effect of aggregate size, buried organic matter, and rhizosphere processes on the occurrence of enhanced N 2 O emissions under simulated flooding/drying conditions in a mesocosm experiment. We used two model soils with equivalent structure and texture, comprising macroaggregates (4000–250 µm) or microaggregates ( 250 µm) from a N-rich floodplain soil. These model soils were either planted with basket willow ( Salix viminalis L.), mixed with leaf litter, or left unamended. After 48 hours of flooding, a period of enhanced N 2 O emissions occurred in all treatments. The unamended model soils with macroaggregates emitted significantly more N 2 O during this period than those with microaggregates. Litter addition modulated the temporal pattern of the N 2 O emission, leading to short-term peaks of high N 2 O fluxes at the beginning of the period of enhanced N 2 O emissions. The presence of S. viminalis strongly suppressed the N 2 O emission from the macroaggregated model soil, masking any aggregate size effect. Integration of the flux data with data on soil bulk density, moisture, redox potential and soil solution composition suggest that macroaggregates provided more favorable conditions for spatially coupled nitrification–denitrification, which are particularly conducive to net N 2 O production, than microaggregates. The local increase in organic carbon in the detritusphere appears to first stimulate N 2 O emissions, but ultimately, respiration of the surplus organic matter shifts the system towards redox conditions where N 2 O reduction to N 2 dominates. Similarly, the low emission rates in the planted soils can be best explained by root exudation of low-molecular weight organic substances supporting complete denitrification in the anoxic zones, but also by the inhibition of denitrification in the zone above, where rhizosphere aeration takes place. Together, our experiments highlight the importance of microhabitat formation in regulating O 2 content and the completeness of denitrification in soils during drying after saturation. Moreover, they will help to better predict the conditions under which hotspots and moments of enhanced N 2 O emissions are most likely to occur in hydrologically dynamic soil systems like floodplain soils.