Bioinformatics analysis
We conducted the de-multiplexing and de-noising of the samples in the QIIME2 (version 2019.10) pipeline (Bolyenet al. 2019) and used the DADA2 method (Callahan et al. 2016) to get rid of artefacts and chimeras. Only amplicon sequence variants (ASVs) that survived the filtering step were kept for subsequent analyses. We trained a new SILVA V4 Classifier (SSU release 132 515-806) by using QIIME2 tutorials as a reference (Quastet al. 2012). ASVs were then assigned a taxonomy with the highest resolution possible (level 7). Following the taxonomic assignment, we split the analysis into two parts: we kept the original output of the taxonomy assignment (unfiltered) and then we filtered the chloroplast and mitochondria assigned reads (filtered). This step was necessary to evaluate the effect of the clamps on the percentage of reads that were allocated to the chloroplasts and mitochondria before and after application of the clamps. The script for our analysis is deposited in Github (https://github.com/luisvqz/V4_pna_clamps_4_wildlife). Further analyses were performed in R [version 3.4.4 (2018)] by using the phyloseq package (McMurdie & Holmes 2013) in a Linux environment. Plots were generated in R by using the package ggplot2(Wickham 2016).