Plant community changes in the arctic: effects on Carbon and Nitrogen stocks distribution in the environment.

Arctic environments undergo important climatic changes that affect, among others, hydrology, soil processes, and plant communities of these systems. At large scale, tree-line and shrub cover have been reported to expand northward, although permafrost melting, increased snow cover and raised soil water content can promote herbaceous covers at the local scale. Our study evaluated carbon (C) and nitrogen (N) stocks in diverse environments at Abisko, northern Sweden: a mire site with palsa, bog, and fen and a shrub tundra site with a bog to broad-leaved forest gradient. Based on plant community survey and vegetation and soil C and N analysis, results showed that proportions of ligneous and herbaceous covers do not reflect the total biomass C and N stocks, with 140.1 ± 56.9 and 3.7 ± 1.5 g per square meter of ground-level vegetation on average, respectively. However, differences in the distribution of short-lived (e.g. leaves) and long-lived (e.g. woody) biomasses were found, with an increase of 1% to up to 40% of woody biomass in dryer sites. Those results were even more important in the broad-leaved forest where C and N stocks in wood, leaves and deadwood of birch trees were over thrice the stock of ground-level vegetation and represented 515.0 ± 115.9 and 17.0 ± 3.8 g.m-2, respectively. Regarding soils, C and N stocks varied mainly at large scale between the mire (47.1 ± 9.1 kgC.m-2 and 2.6 ± 0.4 kgN.m-2 for palsa; 20.2 ± 6.9 kgC.m-2 and 0.9 ± 0.4 kgN.m-2 for bog subsites) and other dryer environments (5.8 ± 1.4 kgC.m-2 and 0.21 ± 0.02 kgN.m-2 for shrub tundra and forest) with differences mostly driven by soil density, soil depth, and water content and not by the composition of the plant community. Our results suggest that plant community shrubification at a large scale is likely to increase the overall C and N stocks in these ecosystems, with more important stocks in long-lived biomass such as wood. While plant community composition and proportion of ligneous/herbaceous species seemed to be a good indicator of biomass distribution, soil stocks appeared not to be well predicted by our results. Those results could be used as a base to compute C and N stocks using remote-sensing data, to obtain information at larger scales for which extensive field measurements are harder to conduct.