2.3. Species richness and composition
In the subsequent analyses, we considered the number of Amplicon Sequence Variants (ASVs) obtained from the ITS as a proxy of the fungal species richness, while the number of ASVs obtained from the 16S were split into Archaea and Bacteria based on the procedure that assigns the Domain. Being interested in deviations from natural conditions of caves, for each cave we calculated the ratio between the ASV richness of each sample collected in the tourist part and the ASV richness of the sample collected in the control area, which is located in the deepest part of the cave, closed to the public (hereafter ASV-ratio).
We then obtained the variation in ASV composition in Fungi, Bacteria and Archaea assemblages by calculating their pairwise dissimilarity among samples by means of the complement of the Sørensen index, which ranges from 0 (samples are composed exactly by the same taxonomic entities) to 1 (samples do not share any taxonomic entity). We used the function ‘beta’ in the BAT package (Cardoso et al., 2015), which allowed us to decompose the total dissimilarity (total β-diversity) into the contribution of its additive components (β-replacement and β-richness), following the partitioning framework independently developed by Podani & Schmera (2011) and Carvalho et al. (2012). The two measures reflect different mechanisms shaping species composition, with β-replacement (βrepl ) reflecting dissimilarity explained by species replacement and β-richness (βrich ) reflecting dissimilarity explained by species loss/gain (richness differences).
For each cave we calculated the dissimilarity measures between each sample collected in the tourist part with respect to the community sampled in the control area. Thanks to this approach, we obtained a measure of the distance of each sample from the control communities.