Table 1.
Description of five study locations.
Fig 1.
Study area locations within the Svalbard archipelago.
Table 2.
GLM and GAM models results of the responses of the tested variables to soil δ15N (response curves given in Fig 2).
Fig 2.
GAM and GLM response curves for increasing soil δ15N values: (A) total soil N content, (B) δ15N of vascular plants, mosses and lichens, (C) total percentage cover of vascular plants and mosses, and Hill’s N2 diversity index, (D) percentage cover of selected plant species (detailed results of the models are given in Table 2, B–D models with scatter plots are presented in the supplementary materials S1 Fig).
Fig 3.
RDA ordination diagram based on N and C contents (%) and isotopic signatures (δ15N and δ13C, ‰) of vascular plants, mosses and lichens (response variables) in relation to the same parameters in soil (explanatory variables).
Table 3.
Total and explainable conditional effects of the environmental variables on vegetation total N and C content, δ15N and δ13C (RDA), and on vascular plant community composition (CCA).
Fig 4.
CCA ordination diagrams based on vascular plant community composition: (A) plant species centroids, with those related significantly positively (blue square) and negatively (yellow square) to soil δ15N (on the basis of t-value biplot); (B) sample plots with pie charts describing the total proportion of positively (blue) or negatively (yellow) related species to soil δ15N in plots. Circle size reflect soil δ15N content and red arrows indicate significant environmental variables.
Fig 5.
Unique and shared fractions of the total variation of vascular plant community composition, explained by geographic location and soil parameters (δ15N, moisture, pH and plot order) (A), with the contribution of each soil parameter to the variation in the soil properties (B).