Soil Nutrients, Landscape Age, and <i>Sphagno-Eriophoretum vaginati</i> Plant Communities in Arctic Moist-Acidic Tundra Landscapes

Abstract

Most research exploring the relationship between soil chemistry and vegetation in Alaskan Arctic tundra landscapes has focused on describing differences in soil elemental concentrations (e.g. C, N and P) of areas with contrasting vegetation types or landscape age. In this work we assess the effect of landscape age on physico-chemical parameters in organic and mineral soils from two long-term research sites in northern Alaska, the Toolik Lake and Imnavait grids. These two sites have contrasting landscape age but similar vegetation composition. We also used correlation analysis to evaluate if differences in any of these parameters were linked with between-site variation in the abundance of growth forms. Our analysis was narrowed to soils in Sphagno-Eriophoretum vaginati plant communities. We found no significant differences between these sites for most parameters evaluated, except for total Ca which was significantly higher in organic soils from Imnavait vs. Toolik and total Na which was significantly higher in mineral horizons from Toolik compared to Imnavait. Moreover, the abundance of non-Sphagnum mosses was positively correlated with total Ca in organic soils, whereas the abundance of forbs, non-Sphagnum mosses and bryophytes was negatively correlated with total Na in mineral soils. We suggest that differences in the concentration of these two elements are most likely tied to landscape age differences between these sites. However, since observed dissimilarity in terms of total Ca in organic soils and total Na in mineral soils is concordant with correlation patterns observed between these elements and the aforementioned growth forms, it is likely that existing differences in vegetation composition between these sites are also influencing the concentration of these elements in soils, particularly that of Ca, since non-Sphagnum mosses are dominant above organic soils and are therefore expected to significantly influence biogeochemical processes at this horizon. Thus, we conclude that except for organic Ca and mineral Na, there is little difference between these sites in terms of their soil physico-chemical properties. We suggest that most of the influence of landscape age on evaluated parameters is masked by factors such as moderate cryoturbation and similarities in terms of vegetation properties and climate. These observations are relevant as they suggest a linkage between soil chemistry and vegetation composition in this tundra region.