Taxonomic allocation of sequence readings and diversity index
The paired-end reading sequences generated from Illumina MiSeq were processed using the software package “Quantitative Insights into Microbial Ecology 2” (QIIME 2, v2018.6) (Caporaso et al. 2010). Briefly, readings were demultiplexed, trimmed, filtered, and merged with the DADA2 complement (Callahan 2016), keeping the sequences with a minimum quality score of 25, a minimum length of 240 bp for reverse readings, and a maximum length of 260 bp for advanced readings. Merged reads collapsed into representative sequences or ASV. Then, ASVs were filtered through the novo chimera using VSEARCH (Rognes et al. 2016). The taxonomy of ASV was assigned at a 99% sequence identity based on the UNITE v7 database (Kõljalg 2013). Non-fungal sequences were removed from the subsequent analysis, and the ASV table was rarefied to a uniform depth (100,000 sequences per sample) to reduce bias related to the depth of sequencing. Taxonomy and shared files produced in QIIME 2 were imported into R (R Core Team 2018) using the Phyloseq package (McMurdie et al. 2013) where α and β diversity were calculated.