Methane and nitrous oxide emissions from South East Queensland waterways

Abstract

Methane (CH4) and nitrous oxide (N2O) are highly potent greenhouse gases (GHGs) with long-term influences on the climate and global warming potentials of 34 and 298 times that of carbon dioxide, respectively. Man-made reservoirs and modified estuaries have been identified as significant aquatic sources of these GHGs. However, there are limited data of the spatial and temporal patterns of CH4 and N2O emissions from these aquatic systems and research on the sources and sinks of these gases are scarce. The lack of data is particularly pronounced for subtropical regions in the Southern Hemisphere. The main aims of this thesis are to investigate CH4 and N2O emissions from a selected subtropical estuary (Brisbane River estuary) and a reservoir (Gold Creek Reservoir) and to identify sources and sinks of these gases in the selected waterways. Methods which facilitate the accurate determination of CH4 and N2O emissions were examined and improved. The accurate determination of liquid and gaseous concentrations of CH4 and N2O is essential to investigate water-air fluxes, sediment-water fluxes and production and consumption processes in the water column of aquatic systems. A systematic assessment of Exetainer vials, which are commonly used as sampling and measurement vials for the determination of liquid and gaseous CH4 and N2O, was conducted. Varying residual air pressure between 0.071 ± 0.008 atm and 0.180 ± 0.031 atm in commercially available, pre-evacuated Exetainer vials was detected leading to potentially substantial background contaminations for emission and liquid phase measurements of CH4 and N2O. Thus, a pre-treatment is suggested which reduces the background of CH4 and N2O concentrations to the best possible result of approximately 3-4% of their respective concentrations in air by flushing vials with nitrogen gas for 5 minutes or alternatively for gaseous samples by flushing with the sampling gas. Methane emissions in the Gold Creek Reservoir were mainly resulting from ebullition (60 to 99%) rather than from diffusive fluxes. Averaged CH4 emissions ranged between 414 and 306,000 µmol CH4 m-2 d-1 among eight sampling sites. Nitrous oxide emissions, ranging between 0.6 and 4.2 µmol N2O m-2 d-1, were primarily driven by diffusion. The findings stressed the importance of monitoring CH4 emissions with appropriate spatial resolution and methods to ensure capture of ebullition zones. In contrast, the assessment of N2O can concentrate on diffusive emissions. A closer investigation at two sites showed that highest dissolved CH4 concentrations were found in the anoxic zones of hypolimnion (600 ± 28 µmol CH4 L-1) and sediment pore waters indicating the sediments as a main source of CH4. Highest dissolved N2O concentrations were found in the epilimnion (0.017 ± 0.001 µmol N2O L-1) and metalimnion (0.023 ± 0.004 µmol N2O L-1), where oxic or oxic/anoxic conditions may have facilitated N2O production by nitrification or denitrification. The Brisbane River estuary was shown to be an overall source of CH4 and N2O with surface water concentrations ranging between 0.048 ± 0.001 and 0.687 ± 0.012 µmol CH4 L-1 and between 0.01 ± 0.003 and 0.02 ± 0.0004 µmol N2O L-1 for CH4 and N2O, respectively, determined at 16 sampling sites. Average CH4 emissions ranged between 136 and 2,603 µmol CH4 m-2 d-1 detected at three selected sites located in the upper, middle and lower reaches. Average N2O emissions ranged between 3.5 and 25.2 µmol N2O m-2 d-1 among the three sites. Large variations in emissions were measured over the tidal cycle at two of the sites in the middle and lower reaches. For CH4, elevated emissions were measured just before or at slack tides. Nitrous oxide emissions varied but no clear changes with the tidal cycle were detected. High tidal cycle and spatial variability highlighted the need for an adequate and comprehensive sampling when measuring estuarine emissions. Of the factors influencing the observed spatial and temporal patterns of CH4 and N2O emissions, tidal currents and surface water concentrations were shown to play an important role. Wind induced turbulence contributed to the emissions, but this did not explain the magnitude and variations in emissions. Furthermore, results from thin boundary layer models, where wind and current speed models are used to determine the gas transfer coefficient (k) were compared to k derived from floating chamber flux measurements. However, k could not be successfully calculated based on six wind and five current speed models alone or in combination at any of the three estuarine sampling sites in comparison to k calculated from the measured emissions. These results stress the importance for direct emission measurements obtained by floating chambers in highly variable systems like the Brisbane River estuary. Sediments were shown to be estuarine sources for CH4 and N2O, whereas the water column was identified as a sink for CH4 and neither a sink nor a source for N2O.