DNA amplification and library preparation

At the molecular laboratory we first discarded the alcohol from the tubes and let the samples dry on large petri dishes with a sterile filter paper inside a fume hood for 3 h. Once dry, we incubated the samples in a lysis buffer (2 mL buffer ASL (Qiagen) + 250 µL proteinase K) for 14 h at 56 °C with a gentle rotation. Subsequently, we recovered the lysate and rinsed the arthropods with distilled water and 90% ethanol, and we stored them in 80% ethanol. We used 225 µL of the lysate to perform DNA extraction on a KingFisher Duo robot (Thermo Fisher Scientific) with the Cell & Tissue kit (Thermo Fisher Scientific), with an elution volume of 100 µL. We included lysis (empty vial with just lysis buffer incubated alongside the samples) and extraction (just for the robot part) blanks.
As metabarcoding markers, we amplified two fragments of the mitochondrial cytochrome oxidase I (COI) and the 16S rRNA genes with the primers COIBF2-COIBR1 (Elbrecht & Leese, 2017) and Chiar16SF-Chiar16SR (Marquina et al., 2019b) respectively. Forward and reverse primers for both markers had attached an 8 bp tag at the 5’ end for sample multiplexing (Binladen et al., 2007) (see Table S4 for the multiplexing information). The PCR reactions consisted of one Illustra Hot Start Taq Mix RTG bead (GE Healthcare Life Sciences), 1 µL of each primer (10 nM), 2 µL DNA template and 21 µL of biology-grade water, for a final volume of 25 µL; the temperature protocol was as follows: initial DNA denaturation and Taq activation at 95 °C for 5 min, 40 cycles of denaturation at 95 °C for 30 s, annealing at 48/50 °C (COI/16S) for 45 s and extension at 68 °C for 45 s, and a final phase of extension at 72 °C for 10 min. We ran all PCRs in duplicated, and a blank was included in the batches of both markers. We confirmed PCR amplification in an agarose gel, pooled the replicates together, and measured DNA concentration using a Qubit 3.0 Fluorometer (Thermo Fisher Scientific) with the broad-range reagents. We discarded lysis, extraction and PCR blanks after not producing a visible band in the gels and returning undetectable values of DNA concentration. We then pooled the tagged PCR products equimolarly, and cleaned-up primers and primer dimers using MinElute columns (Qiagen). We prepared libraries with the TrueSeq PCR-free kit (Illumina). We visualized ligation products on a 2% agarose gel, cut the fragments of the desired length out of the gel, and purified them using the QIAquick gel extraction kit (Qiagen). We pooled the libraries to equimolar concentrations and sequenced them on a single Illumina MiSeq lane using 2 x 300 bp paired-end v3 chemistry run at SciLifeLab (National Genomics Infrastructure, Stockholm).