The denitrification process and the associated nitrous oxide (N(2)O) production in soils have been poorly documented, especially in terms of soil profiles; most work on denitrification has concentrated on the upper layer (first 20 cm). The objectives of this study were to examine the origin of N(2)O emission and the effects of in situ controlling factors on soil denitrification and N(2)O production, also allowing the (N(2)O production)/(NO(3) (-)-N reduction) ratio to be determined through (1) the position on a slope reaching a river and (2) the depth (soil horizons: 10-30 and 90-110 cm). In 2009 and 2010, slurry batch experiments combined with molecular investigations of bacterial communities were conducted in a corn field and an adjacent riparian buffer strip. Denitrification rates, ranging from 0.30 mu g NO(3) (-)-N g(-1) dry soil h(-1) to 1.44 mu g NO(3) (-)-N g(-1) dry soil h(-1), showed no significant variation along the slope and depth. N(2)O production assessed simultaneously differed considerably over the depth and ranged from 0.4 ng N(2)O-N g(-1) dry soil h(-1) in subsoils (the 90-110-cm layer) to 155.1 ng N(2)O-N g(-1) dry soil h(-1) in the topsoils (the 10-30-cm layer). In the topsoils, N(2)O-N production accounted for 8.5-48.0% of the total denitrified NO(3) (-)-N, but for less than 1% in the subsoils. Similarly, N(2)O-consuming bacterial communities from the subsoils greatly differed from those of the topsoils, as revealed by their nosZ DGGE fingerprints. High N(2)O-SPPR (nitrous oxide semi potential production rates) in comparison to NO(3)-SPDR (nitrate semi potential reduction rates) for the topsoils indicated significant potential greenhouse N(2)O gas production, whereas lower horizons could play a role in fully removing nitrate into inert atmospheric N(2). In terms of landscape management, these results call for caution in rehabilitating or constructing buffer zones for agricultural nitrate removal.
