The effect of vegetation type and snow depth on annual CO2 fluxes in a high arctic tundra region

Global circulation models predict an increase in precipitation in the Arctic, especially during winter time. It is likely that alterations of the snow cover, which has an influence on gas exchange between the soil and the atmosphere, will also affect annual gas-fluxes as arctic regions are snow-covered most of the year. This project aims to characterize the effect of vegetation (type and amount of cover) on annual CO₂ fluxes and investigates the effect of increased snow depth on CO₂ fluxes and soil temperature in an arctic tundra region. Twelve snow fences have been set up to manipulate the natural snow accumulation. These fences were established in Adventdalen, Svalbard (78°N, 16°E) in autumn 2006 and are 1.5 m high, 5 m long and placed in four different study areas. The study areas differ in plant composition and soil moisture conditions; two are on arctic heath (dominated by Cassiope tetragona), two are on arctic meadow (dominated by Luzula arcuata ssp. confusa). Measurements of CO₂ fluxes started in July 2007, continued through the autumn and winter and will proceed until end of summer 2008. A portable LiCor infrared gas analyzer (LiCor 6400-09/6262 Soil CO₂ Flux Chamber, LiCor, Lincoln, USA) with a closed dark soil-flux chamber is used to measure CO₂ efflux on collars installed at the soil surface.
Preliminary results indicate that CO₂ efflux and soil temperature differed between the two vegetation types. Effluxes were generally higher on arctic meadow which had higher soil temperatures than arctic heath. The difference in CO₂ effluxes and soil temperatures between the two vegetation types reduced with the onset of winter. Effluxes generally decreased towards the winter when CO₂ efflux rates were about 3-16% of those measured in the summer. Snow fences increased the snow to a maximum depth of 1.7m compared to 10-30cm in control areas. Soil temperatures behind the fences were increased by 2-10°C in winter and decreased by a similar amount in early summer. The decrease of CO₂ efflux towards the winter was smaller on sites with snow fences and the efflux in winter was about 55% higher in these areas compared to the controls. Broadly speaking the difference in measured CO₂ efflux for the different treatments was more pronounced in winter time. Results so far indicate that there are differences in CO₂ efflux between different vegetation types and that an increase in snow depth can lead to increased CO₂ efflux during winter time, possibly affecting the carbon budget of arctic tundra regions.