Ecosystem CO₂ production during winter in a Swedish subarctic region: the relative importance of climate and vegetation type

General circulation models consistently predict that regional warming will be most rapid in the Arctic, that this warming will be predominantly in the winter season, and that it will often be accompanied by increasing snowfall. Paradoxically, despite the strong cold season emphasis in these predictions, we know relatively little about the plot and landscape-level controls on tundra biogeochemical cycling in wintertime as compared to summertime. We investigated the relative influence of vegetation type and climate on CO₂ production rates and total wintertime CO₂ release in the Scandinavian subarctic. Ecosystem respiration rates and a wide range of associated environmental and substrate pool size variables were measured in the two most common vegetation types of the region (birch understorey and heath tundra) at four paired sites along a 50 km transect through a strong snow depth gradient in northern Sweden. Both climate and vegetation type were strong interactive controls on ecosystem CO₂ production rates during winter. Of all variables tested, soil temperature explained by far the largest amount of variation in respiration rates (41-75%). Our results indicate that vegetation type only exerted an influence on respiration when snow depth was below a certain threshold (~1 m). Thus, tall vegetation that enhanced snow accumulation within that threshold resulted in more effective thermal insulation from severe air temperatures, thereby significantly increasing respiratory activity. At the end of winter, within several days of snowmelt, gross ecosystem photosynthesis rates were of a similar magnitude to ecosystem respiration, resulting in significant net carbon gain in some instances. Finally, climate and vegetation type were also strong interactive controls on total wintertime respiration, suggesting that spatial variations in maximum snowdepth may be a primary determinant of regional patterns of wintertime CO₂ release. Together, our results have important implications for predictions of how the distribution of tundra vegetation types and the carbon balances of arctic ecosystems will respond to climate change during winter because they indicate a threshold (~1 m) above which there would be little effect of increased snow accumulation on wintertime biogeochemical cycling.