Comparison of analytical procedures to estimate CH4 and N2O fluxes from a two-year ecosystem study in a constructed wetland system

Abstract:

Constructed wetland systems (CWS) have been developed to remove nutrients from secondarily treated water, but little is known about their long-term contributions on greenhouse gas emissions (GHG), especially in arid regions. To increase our knowledge of ecosystem dynamics of CWS in arid regions, we are investigating N₂O, CH₄, and CO₂ fluxes from a system perspective, a vegetated-shoreline to open-water gradient, and from the wetland plant Typha spp. From 2012 to 2014, we utilized the floating chamber technique to collect fluxes from two transects (nearest to inflow and nearest to outflow) and along two gradient subsites (shoreline and open-water) within the transects. Recently, we began collecting direct fluxes from the vegetation by deploying gas chambers on Typha spp. Fluxes were analyzed using the HMR procedure (Package HMR in R) developed for trace-gas flux estimations when using static chambers. We found significantly higher CH₄ and CO₂ fluxes in the summer and spring compared to fall and winter months. From the whole system perspective, we found significantly greater CO₂ fluxes at the inflow compared to the outflow transect. From the shoreline to open-water gradient, N₂O fluxes were significantly greater in the open-water and, CH₄ fluxes where significantly greater in the vegetated shoreline subsite. These differences may be explained by the presence of vegetation, differences of water column height, or higher nitrate levels in the open-water compared to the shoreline. Results from the vegetation chambers will be presented from two heights of the Typha spp. leaves from plants in each of the four subsites. The analysis of the 288 fluxes using two HMR procedures, default classification (linear, non-linear, and no flux) and the linear regression, resulted in similar seasonal and spatial patterns in the flux estimates. However, the default classification calculated on average 31% for N₂O, 67% for CH₄, and 34% for CO₂ higher flux estimates relative to the fluxes estimated only by linear regression. Our results highlight the importance of weather conditions, vegetation, and nutrient loading on GHG fluxes in a CWS. Finally, it is important to establish long-term ecosystem-wide studies and at the same time, maintain a constant evaluation of analytical methods available for calculating GHG flux estimates.