Molecular analyses
Soil DNA was extracted using MoBio PowerSoil kits ® following manufacturer’s instructions. Prokaryotic and fungal DNA were amplified targeting the 16S and ITS rDNA regions, respectively. The primers used for the amplification were the 515f-806R couple (targeting the hypervariable V4 region) for 16S and the ITS1f-ITS2 couple for ITS (targeting the ITS1 subregion) (White et al. , 1990; Gardes & Bruns, 1993; Gilbert et al. , 2014). PCR amplifications were conducted with 0.2 μM primers, 0.2mM dNTPs, 0.5 units of hi-fidelity Taq polymerase (New England Biolabs), and 1 μL of DNA extract in 25 μL reactions. Thermocycling was done as follows: initial denaturation at 95°C for 1 minute, 30 cycles of denaturation (95°C 30s) - annealing (16S: 50°C / ITS: 52°C for 30s) – elongation (72°C 1 min), and a final, 5 min extension at 72°C. Blanks were included to ensure absence of DNA amplification when no soil was added in DNA extraction kits. Positive amplicons were purified and sequenced using the Illumina MiSeq technology (Genome Québec Innovation Center).
Sequences were demultiplexed and paired-ends were merged in QIIME (Caporaso et al. , 2010), after which we proceeded to a quality filtering (quality score > 25) and a removal of sequences with homopolymers longer than 10 bases. Chimera removal and operational taxonomical unit (OTU) clustering was performed in the open source software VSEARCH (Rognes et al. , 2016), at a 97% similarity threshold. Taxonomy was assigned in QIIME using the curated databases Silva and UNITE (Pruesse et al. , 2007; Abarenkov et al. , 2010). We removed OTUs occurring in only one sample and occurrences with 5 reads or less, considering these as potential artefacts. We rarefied our OTU tables (16S and ITS) to 8000 reads per sample for fungi and 1100 reads for prokaryotes to correct for imbalanced sequencing depth per sample, due to the stochastic nature of sequencing. Sequence data were deposited in SRA under project # 7716646.