Greenhouse gas fluxes monitored for 5 years in two drained boreal peatlands using eddy covariance and automatic light-dark chambers

Peatlands store large amounts of organic carbon, which become subject to increased microbial decomposition and mineralization to primarily CO2 upon drainage. Drained peatlands are often characterized by horizontal variability in soil water content and saturation, with drier parts closer to drainage ditches. This causes spatial heterogeneity in greenhouse gas fluxes, where CH4 emissions are higher in the wetter areas, while respiratory CO2 production dominates in drier areas. We investigated two neighboring, drained ombrotrophic bogs in Norway close to Trysil, Innlandet, 61.1N- 12.25E, 640 m a. s. l. We used a total of 10 automated, ecosystem-level, light-dark chambers (ECOFluX) to examine the seasonality of CO2, CH4 and N2O fluxes at different microsites along the water table gradient from the center of drained patches to the drainage ditches to relate GHG fluxes to small scale spatial heterogeneity. With eddy covariance measurements of CO2 and CH4 fluxes, we estimated GHG fluxes over representative areas of each bog. In year 3, the natural hydrology of one of the sites (South) was restored while the other site (North) was kept as a drained reference site. We here present a comparative analysis of the five years of measurements, where we examine shifting spatial patterns of GHG production at different scales and relate them to soil water conditions. The automated chambers (five chambers within each footprint of each eddy flux tower) showed higher spatial variability for CH4 fluxes than for CO2 with higher CH4 emissions in the wetter plots furthest away from ditches, i.e. CH4 fluxes correlated well with the water table depth at both sites. Very small N2O emissions were observed only during very short events in the early summer of year 1. While the CO2 fluxes were highly comparable at the two investigated sites, the CH4 fluxes were moderately higher at the lower and wetter of the two sites (North). At both sites, the CH4 fluxes correlated well with local variations in groundwater table and overall there was a good alignment of fluxes measured with eddy flux and chamber technologies. Restoring the south site resulted (on average) in 6-10 cm elevation of the ground water table, but in an increase of CH4 emissions of approx. 70%. At the restored site, we observed a decrease in ecosystem respiration and photosynthesis rates compared to the reference site during the first year after restoration. Finally, we will present the overall conclusions of the 5 year study and our recommendations for the stakeholders.