Fungal taxonomic composition
The C1 and C2 samples of control plots were not saturated in sequence depths (less than 10,000 sequences, based on the rarefaction curves in Fig. S1), probably due to technical problems of DNA sequencing. Therefore, they were discarded for data analyses. After resampling of raw 28S rRNA gene reads, a total of 126,050 high-quality sequence reads were generated and were clustered into 991 OTUs at the 97% identity.Ascomycota (70.66±11.80%) was the dominant phylum, followed by Basidiomycota(14.21±10.19%), Chytridiomycota (0.51±0.64%) andBlastocladiomycota (< 0.01%). Those OTUs were assigned to 135 genera and the most abundant genera included Kriegeria(9.14±9.64%), Hyaloscypha (6.63±5.41%), Collophora(3.43±6.50%), Mollisia (2.20±3.68%), Russula(0.80±3.40%), Sorocybe (0.57±1.77%), Rigidoporus(0.46±1.53%), Chloroscypha (0.45±0.78%), Hymenoscyphus(0.41±0.61%) and Crinula (0.40±1.00%). The fungal genera were further classified into 6 ecological groups based on their roles in ecosystems (Tedersoo et al., 2014) - saprotrophs (57%), plant pathogens (24%), ectomycorrhizal fungi (6%), animal parasites (5%), mycoparasites (2%) and mycobionts (6%), among a total of the 135 genera. However, soil warming did not change the overall fungal community composition, reputing part of our first hypothesis (Table 1 & Fig. S2a). Additionally, no differences in diversity and environmental group abundances were found between warmed and control samples (Table S3).