Library preparation and sequencing
For sequencing, we amplified a short hypervariable fragment of the 18S SSU rRNA V7 region (~100 – 110 bp) with 18S_allshorts primers (Forward 5’-TTTGTCTGSTTAATTSCG-3’, and Reverse 5’-GCAATAACAGGTCTGTG-3’) (Guardiola et al., 2015). Broad coverage within Eukarya was verified with the Arb Silva TestPrime function (Klindworth et al., 2013). To enable pooling of samples after amplification, primers were pre-tagged with an 8-base oligonucleotide, wherein at least 3 nucleotides differed between tags. The primers also contained a leading 5’-end variable number of degenerate nucleotides (N, 2 ≤ 4) to increase sequence variability and hence Illumina sequencing quality (Wangensteen et al., 2018). PCR amplification and all downstream processing from here and onwards were conducted twice. First as a pilot with a limited number of samples and negatives (N­­1 = 79), and finally with a full set of samples and negatives (N2 = 456). Samples were processed using the exact same protocol unless stated otherwise.
PCR amplifications were performed in 20 µL reactions with 10.00 µL AmpliTaq Gold ™ Master Mix (Applied Biosystems), 0.16 µL Bovine Serum Albumin (BSA, 20 µg µL-1), 5.84 µL ultrapure MQ water, 2.00 µL of 18S_allshorts Forward and Reverse primer mix (2.5 µM each), and 2.00 µL DNA template. Thermal cycling consisted of an initial denaturation step (10 min, 95°C), and 35 cycles of denaturation (30 s, 95°C), annealing (30 s, 45°C) and elongation (30 s, 72°C). For each PCR plate, a subsample of real samples (n = 9) and a PCR negative were tested on a 1% agarose gel to verify amplification and a lack thereof, respectively. PCR amplicons were then pooled, purified with MinElute ™ spin-columns (Qiagen, Hilden, Germany) and quantified using the broad-range dsDNA assay on a Qubit 4™ fluorometer (Invitrogen by ThermoFisher). Sequencing-ready libraries were prepared from purified pools in accordance with the NEXTflex™ PCR-Free DNA Sequencing Kit (Bioo Scientific, Austin, Texas, USA). Here, DNA templates (3000 ng total for each library) were first purified by size (retaining fragments >150 bp) with magnetic Agencourt AMPure XP beads (Beckman Coulter Genomics, California, USA), then adenylated, and ligated with Illumina-compatible adapters. We used several Illumina-compatible adapters (NEXTflex™ DNA Barcode Adapter, Bioo Scientific) to distinguish libraries of 96 samples. The libraries were quantified by qPCR using the NEBNext® Library Quant Kit for Illumina® (New England Biolabs, Massachusetts, USA) to verify successful preparation. For the pilot sequencing one library of 79 samples (75 real and 4 extraction negatives), were sequenced using 150 Paired-End (PE) chemistry on a HiSeq 4000 platform (Novogene Co., Ltd.). For the full sequencing run, five libraries with a total of 456 samples (437 real and 19 extraction negatives), were sequenced using 150 PE chemistry on two lanes of a NovaSeq6000 platform (Novogene Co., Ltd.).