Carbon Fluxes in a sub-arctic tundra undergoing permafrost degradation

Wednesday, 17 December 2014
Rosvel G Bracho1, Elizabeth Webb1, Marguerite Mauritz2 and Edward A G Schuur1, (1)University of Florida, Gainesville, FL, United States, (2)University of Florida, Ft Walton Beach, FL, United States
As an effect of climate change, temperatures in high latitude regions are increasing faster than in the rest of the world and future projections indicate it will increase between 7°C and 8°C by the end of the 21st century. Permafrost soils store around 1700 Pg of Carbon (C), which is approximately the amount of C stored in terrestrial vegetation and in the atmosphere combined. Sustained warming induces permafrost thaw, leads to a thicker seasonal active layer, and creates subsided patches in the landscape. Carbon that was previously inaccessible to decomposition is thus exposed, increasing the likelihood of positive feedback of CO2 to the atmosphere. We measured C fluxes (Net ecosystem carbon flux, NEE, and Ecosystem respiration, Re) using the eddy covariance approach in a tundra landscape (Eight Mile Lake Watershed, Alaska) undergoing permafrost degradation from the beginning of the growing season in 2008 and throughout most winters until May 2014. This interval encompassed a range of climatic variability that included a deviation of ± 50% from the long term average in growing season precipitation. Active layer depth (thaw depth at the end of the growing season) and subsidence in the footprint were used as indicators of permafrost degradation. Results indicate that annual NEE ranged from a sink of 0.76 MgC ha-1 yr-1 to a source of 0.55 MgC ha-1 yr-1. NEE during the growing seasons fluctuated from 1.1 to 1.8 MgC ha-1 season-1 in net C uptake. Annual NEE was strongly affected by winter Re, which represented between 33% and 45% of the annual value regardless of of the large drop in both air and soil temperature. Parameters from the light response curve (optimum NEE, NEEopt and quantum yield, α) showed a seasonal and interannual variability and were different between the most and least degraded sites in the footprint, which affected the magnitude of the carbon cycle and may have implications for landscape C balance in sub-arctic tundra.