P inputs determine denitrifier abundance explaining dissolved nitrous oxide in reservoirs

Reservoirs are important sites for nitrogen processing, especially those located in agricultural and urban watersheds. Nitrogen inputs promote N2O production and emission, but the microbial pathways controlling N2O have been seldom studied in reservoir water columns. We determined N2O concentration in the water column of 12 reservoirs during the summer stratification and winter mixing. We explored the potential microbial sources and sinks of N2O by quantifying key genes involved in ammonia oxidation (bacterial and archaeal amoA) and denitrification (nirS and nosZ). Dissolved N2O varied up to three orders of magnitude (4.7-2441.2 nmol L-1) across systems, from undersaturated to supersaturated values (37%-24,174%) depending on reservoirs and depths. N2O concentration depended on nitrogen and oxygen availabilities, with the lowest and highest N2O values at suboxic conditions. Ammonia-oxidizing archaea dominated over ammonia-oxidizing bacteria but were not related to the dissolved N2O. In contrast, the abundance of the nirS gene was significantly related to N2O concentration, and three orders of magnitude higher than amoA abundance. Denitrifying bacteria appeared consistently in the water column of all reservoirs. The nirS and nosZ genes appeared in oxic and suboxic waters, but they were more abundant in suboxic waters. The nitrate concentration, and nirS and nosZ relative abundances explained the dissolved N2O. Besides, nirS abundance was related positively with total phosphorus and cumulative chlorophyll a, a proxy for fresh organic matter. Therefore, P inputs, not just N inputs, promoted N2O production by denitrification in the water column of reservoirs.