Nitrous oxide yields from urban stormwater ponds in cities across the US

Background/Question/Methods:

Numerous stormwater retention ponds are built in US cities to manage storm flows; however, their effect on biogeochemical cycles, specifically denitrification, are unknown. The high temperatures, low oxygen and high nitrate typical of these urban water bodies should provide ideal conditions for denitrifying microbial communities. As a result, stormwater ponds may be major sources of both N₂O and N₂ fluxes within urban landscapes. Previous work in urban aquatic sediments has also suggested that stress from common trace metal contaminants (i.e. Zn, Cd, Pb) might shift the denitrifying microbial community or efficiency in a way that reduces the importance of the N₂O reduction step, thereby increasing the relative amount of nitrous oxide produced during denitrification. Here we ask the question: Which conditions in pond sediments in 8 US cities best explain the variation observed in N₂O and N₂ production in urban ponds relative to their green space counterparts? To address this question, during the summer of 2014 we collected sediments from 4 highly urban and 4 ‘green space’ ponds within 8 USA metropolitan statistical areas (MSAs) including: Durham, NC; Phoenix, AZ; Miami, FL; Baltimore, MD; Boston, MA; St. Paul, MN; Salt Lake City, UT; and Portland, OR. We measured potential N₂O and N₂ production, porewater chemistry, sediment chemistry, and heavy metal concentrations for sediments from each pond.

Results/Conclusions:

Our findings show a tremendous amount of variability in N₂O and N₂ potential production rates both within and across cities, with N₂O rates ranging from 0.002 to 37.6 mg N₂O kg^-1 hr^-1 and N₂ rates ranging from 0 to 44.2 mg N₂ kg^-1 hr^-1. N₂O accounted for as little as 0.01% to as much as 99% of total gaseous N fluxes and there were no consistent differences in N fluxes between high density and green space ponds. Substrate supply (i.e. labile C and nitrate) explained less than half of the variability in total N flux (R²=0.24; p<0.005). We found a preliminary positive relationship between sediment zinc concentrations and N₂O rates (R²=0.25; p<0.05), but no relationship between Zn and N₂ production suggesting some interaction between contaminant stress and the importance of the N₂O reduction step. We are currently exploring other factors that might influence denitrification such as disturbance by hydrologic variability quantified through landscape analysis. Our findings suggest that the relative importance of different drivers of N₂O and N₂ production in ponds might be altered by conditions resulting from watershed urbanization.