Figure legends
Figure 1. (a) The nine sampled sites in south-central Norway, from south to north: Haglebu, Ustevatn (Hardangervidda), Strandavatnet, Skyrvedalen (Hemsedal), Storlifjell (Vestre Slidre), Bessheim (Jotunheimen), Lemonsjøen (Vågå), Sel, and Dombås. The map was drawn in QGIS version 3.4.14, with map data from geonorge.no. (b) Photo of the abrupt vegetation change at the forest line at the Lemonsjøen site. (c) Schematic view of the studied transects stretching across the mountain birch forest line. Numbers 1 to 11 represents the plots where soil samples were collected every 20 meters along the 200 m transect stretching from 100 meters below to 100 meters above the mountain birch forest line.
Figure 2. Barplots illustrating the relative abundances of taxonomic groups across the ecotone based on the number of reads. (a) Kingdom and (b) phylum level distribution of the 18S data. (c) Class level distribution of the fungal 18S data. (d) Distribution of fungal classes based on the ITS2 dataset. Plot number 01 is the lowermost plot in the boreal mountain birch forest, while plot number 11 is the uppermost in the low-alpine vegetation.
Figure 3. Boxplots illustrating how (a) the tree dominant fungal guilds, (b) soil edaphic factors and ergosterol, and (c) plant groups based on mycorrhizal status vary across the ecotone. The bold lines in the boxes show the median, and the upper and lower part of the box represents the 25th and 75th percentiles. Outliers are represented by black dots. Plot number 01 is the lowermost plot in the boreal mountain birch forest, while plot number 11 is the uppermost in the low-alpine vegetation.
Figure 4. Scatter plots showing the relationship between (a) ergosterol and soil carbon content and (b) proportional abundance of dark septate endophytes (number of reads) and soil carbon content. The relationships were modelled using mixed models with sites as a random factor, from which the R squared values shown were extracted. The lines represent least squares regression lines.
Figure 5. GNMDS ordination plots of (a) the 18S dataset and (b) ITS2 dataset based on rarefied OTU matrices. Each point represents one soil sample. Plots belonging to the low-alpine vegetation are colored red/yellow, while plots belonging to the boreal mountain birch forest are green (see legend for details). Variables that significantly correlated with the community structure (p<0.05) are shown as arrows in the diagrams. Dark arrows represent edaphic factors, dark green arrows site-specific factors, while light green arrows illustrates the distribution of mycorrhizal plant types. (c) GNMDS ordination plot of the ITS2 dataset based on species scores. Each point represents one OTU. Only the 40 most common OTUs are visualized. Isolines are numbered according to the plot-numbers ranging from 01 (low-alpine vegetation) to 11 (boreal mountain birch forest). Colors indicate functional groups (see legend), and size of circles are proportional to the number of reads. (1) Pezoloma ericae , (2) Lecanorales sp., (3)Helotiales sp., (4) Cortinarius caperatus , (5)Piloderma sp., (6) Chaetothyriales sp., (7)Cenococcum sp., (8) Chaetothyriales sp., (9)Cortinarius armillatus , (10) Mortierella humilis , (11)Herpotrichiellaceae sp., (12) Cortinarius pseudocandelaris , (13) Cortinarius alpinus , (14),Venturiales sp., (15) Herpotrichiellaceae sp., (16)Pseudotomentella tristis , (17) Meliniomyces sp ., (18)Amanita sp., (19) Helotiales sp., (20)Herpotrichiellaceae sp., (21) Solicoccozyma terricola , (22) Piloderma bicolor , (23) Phialocephala sp., (24)Helotiales sp., (25) Luellia sp., (26)Chaetothyriales sp., (27) Hygrophorus albicastaneus , (28)Mucor abundans , (29) Agaricales sp., (30)Helotiales sp., (31) Meliniomyces sp. 2, (32)Meliniomyces sp. 3, (33) Herpotrichiellaceae sp., (34)Meliniomyces sp. 4, (35) Hyaloscypha sp., (36)Leohumicola minima , (37) Hygrophorus sp., (38)Meliniomyces bicolor , (39) Muriformistrickeria rosae , (40)Cortinarius talus .