Data and scripts for "The importance of within-log sampling replication in bark- and wood-inhabiting fungal metabarcoding studies"

Data and scripts for reproducing the analyses of Naranjo-Orrico et al, "The importance of within-log sampling replication in bark- and wood-inhabiting fungal metabarcoding studies". The input data consists of the following four files, "Alldata.Rdata", "data_SbVenn_meta&morpho.Rdata" , "Xmorpho.csv" and "Ymorpho_1.csv". The former two files are in R format and the latter two in CSV format. The R files need to be loaded using the function load, and the CSV files with the function read.csv2 in R. "Alldata.Rdata" includes in total 15 input data matrices: - Metadata for dataset A (meta22) - Metadata for dataset B (meta23) - Two sample x OTU tables for dataset A including the number of reads for each OTU (otu.table.plausible.2022 for the plausible OTU taxonomic identifications and otu.table.reliable.2022 for the reliable taxonomic identifications). - Two sample x OTU tables for dataset B including the number of reads for each OTU (otu.table.plausible.2023 for the plausible OTU taxonomic identifications and otu.table.reliable.2023 for the reliable OTU taxonomic identifications). - Two sample x OTU tables for dataset A including the relative read counts per OTU (otu.table.plausible.w.2022 and otu.table.reliable.w.2022). - Two sample x OTU tables for dataset B including the relative read counts per OTU (otu.table.plausible.w.2023 and otu.table.reliable.w.2023). - Read counts per sample during the different phases of the bioinformatics pipeline for dataset A (read.counts.plausible.2022) and for dataset B (read.counts.plausible.2023). - Taxonomic information at all taxonomic levels (i.e., form species to phylum) of the identified OTUs (taxonomy.plausible) - Guild assignments matrices for dataset A (Guilds_plausible_tax_2022) and for dataset B Guilds_plausible_tax_2023). "data_SbVenn_meta&morpho.Rdata" contains four matrices: - Occurrence of the lichenized OTUs identified through metabarcoding including identifications at any taxonomic level (i.e., genus or family levels when species level identifications were not achieved) (SbVenn_Lmeta). - Occurrence of the lichenized OTUs identified through metabarcoding including identifications at the species-only level (SB_Venn_clean_meta). - Occurrence of the morphologically identified lichenized fungi, including identifications at the genus level and morphospecies (SbVenn_Lmorpho) - Occurrences of the morphologically identified lichenized fungi, including identifications at the species-only level (SbVenn_clean_morpho). "Xmorpho.csv" and "Ymorpho_1.csv" contain respectively the metadata and the presence-absence data of the morphologically identified lichens. “Alldata.Rdata” is used in all the scripts, "data_SbVenn_meta&morpho.Rdata" is only needed for the script "S8_Venn Diagrams.R", and the the files "Xmorpho.csv" and "Ymorpho_1.csv" are used in "S11_Meta vs Morpho species richnes between tree sp and tree part.R". The statistical analyses consist of joint species distribution modelling with the package Hmsc, generalized linear mixed models (GLMM) with the package glmer, and non-metric multidimensional scaling analysis (NMDS) with the package vegan. To perform the HMSC analyses, the first FOUR scripts need to be run consecutively from S1 (A and B) to S3. S1A defines the first model using data A, and S1B defines the second model using dataset B. S2 fits the models fitted in the study (which include presence-absence with a different set of explanatory variables). S3 shows the parameter estimates from the fitted models, in particular the beta parameters and the variance partitioning across environmental covariates. For fitting and showing the outputs of the GLMM models only S4 is needed. In S5, runs the NMDS analyses. The rest of the scripts, S6-S11 are used to produce the different plots shown in the study of Naranjo-Orrico et al., including pieplots, boxplots, barplots, and Vennplots.