Assessing the Spatial and Temporal Variability of Diffusive Methane and Nitrous Oxide Emissions from Subtropical Freshwater Reservoirs

Surface watermethane (CH4) and nitrous oxide (N2O) concentrations were measured and diffusive fluxes were estimated in three subtropical freshwater reservoirs (Little Nerang Dam (LND), Lake Wivenhoe (LW) and Lake Baroon (LB)) in southeast Queensland, Australia, during four seasons in 2011-2012. All reservoirs were strong sources of CH4 in all seasons. Surface water CH4 varied between 1350 and 524 000% saturation, and was overall highest in spring and summer, and lowest in winter, however, with no clear patterns common to all reservoirs. In contrast, all reservoirs switched from weak N2O sinks in spring to strong N2O sources for the rest of the year. N2O saturation in all reservoirs varied between 70 and 12-30%. There were significant differences for CH4 concentrations and fluxes between the reservoirs. Within each reservoir, there was strong spatial CH4 variability but minimal N2O saturation variability. CH4 saturation was higher in inflow zones than in the main body. Area-weighted average fluxes were estimated using six waterair gas transfer velocity estimation models and resulted in fluxes in the range 4.820.5, 2.35.4, and 2.37.5 mg CH4 m(-2) d(-1), while N2O was 0.070.41, 0.090.22, and 0.030.09 mg N2O m(-2) d(-1) for LND, LW, and LB, respectively. Total emissions, in carbon dioxide equivalents, from all measurement campaigns were CH4 dominated (67-86%). The measured degree of CH4 saturation and fluxes are among the highest reported thus far indicating that subtropical freshwater reservoirs could be significant aquatic greenhouse gas sources. This paper provides a comprehensive assessment of the interplay between biogeochemical processes and the physical forcing driving the waterair gaseous emissions. The high variability coupled with the lack of consensus among estimation models calls for concerted efforts to address uncertainty of measurements for reliable emissions accounting.