Additional file 1: Table S1. Summary of environmental factors. Table S2. Major topological properties of the empirical pMENs of bacterial and fungal communities in the control and warming sites and the associated random networks. Figure S1. Microbial taxonomic composition of (a) bacterial communities and (b) fungal communities at the phylum level (for bacterial communities, Proteobacterial Classes are juxtaposed with other Phyla). Phyla with abundance less than 1% were combined to Others. Figure S2. Networks among environmental factors and microbial communities. (a) Bacterial communities from warming plots; (b) Bacterial communities from control plots; (c) fungal communities from warming plots, and (d) fungal communities from control plots. Red nodes represent environmental factors, blue nodes represent OTUs directly connected to environmental factors, grey nodes represent OTUs indirectly connected to environmental factors. Grey edges represent positive correlations, and red edges represent negative correlations. Abbreviations: Plant, aboveground plant biomass; Moisture, soil moisture; Bulk density, bulk soil density; C, soil total carbon; N, soil total nitrogen, and C/N, soil carbon/nitrogen ratio. Figure S3. The linear regressions between pairwise similarities of (a) bacterial community composition and functional structure (GeoChip data), and (b) fungal community composition and functional structure. Bray-Curtis distance was used for the similarity calculations. Figure S4. Differences of average normalized signal intensities of representative genes involved in (a) nitrogen cycling, (b) phosphorus cycling, and (c) sulfur cycling. (a) Red gene names represent genes with a higher average normalized signal intensity in warming samples, whose percentages of changes are indicated in parentheses. (b & c) Blue bars represent the average normalized signal intensity of gene probes in control samples, and red bars represent the average normalized signal intensity of gene probes in warming samples. Error bars represent standard errors. The differences between warming and control samples were tested using ANOVA, with * indicating p < 0.050. Figure S5. Overall community stochasticity on the basis of phylogenetic metric of (a) bacterial communities and (b) fungal communities. The data for each bar contains n = 15 within-group pairwise comparisons calculated from 6 biological replicates. Figure S6. Relationship between microbial community composition or functional structure and environmental factors revealed by canonical correspondence analysis (CCA) of (a) bacterial communities (red dots represent warming samples, and blue dots represent control samples) and environmental variables (arrows), (b) fungal communities and environmental variables; and (c) microbial functional structure and environmental variables. All CCA models are significant (p < 0.050).
|Date made available||Jan 17 2020|
|Publisher||figshare Academic Research System|