2.1. Sampling design
We adopted a replicated two-factorial sampling design wherein sediment samples were repeatedly sampled at progressive distance from the tourist path and at progressive distance from the entrance in four Italian show caves (Fig. 1).
In each cave, we identified four sampling areas at progressive distance from the cave entrance: Section I to IV, with Section I being the closest to the entrance and Section IV the furthest. Sections I, II and III were located in areas open to the public (touristic areas, T), while Section IV was located in an area closed to the public in the deepest part of the cave (control area, C). The length of each Section was one third of the tourist path length. Within each sampling section of the touristic part, we identified three sampling transects placed at progressive distance from the tourist path: the High Pressure (HP) transect at 0-3 m from the tourist path; the Medium Pressure (MP) transect at 3 to 5 m from the tourist path; and the Low Pressure (LP) transect at > 5 m from the tourist path. In the control area (Section IV), three sampling transects were randomly placed to obtain a representative view of the natural conditions of the cave.
Along each transect, we collected 9 random replicates of alluvial sediment within 9 squares of 10 cm2. Sediment was collected on the ground up to 3 cm depth with sterile Falcon® tubes (50 mL). Samples were preserved in a thermal bag until the arrival at the laboratory, where the 9 replicates for each sub-transect were pooled together and homogenised. In this way, we obtained a total of 12 samples (4 transects × 3 sub-transects) for each show cave. We focused on soil as it proved to be the compartment with the highest microbial diversity (see Alonso et al., 2019). As seasonal differences have been observed in the subterranean microbiota (Mammola et al., 2017), samplings were performed in Summer, between June and September 2020, in order to avoid possible seasonal variations that could mask the effect of tourism on the soil microbiota.
Sediment samples were sieved, under sterile conditions, by removing coarse rock debris. Their physical and chemical properties were evaluated by measuring pH, concentration of Organic Carbon, total Nitrogen and the percentage of sand (%Sand), silt (%Silt) and clay (%Clay) with standard protocols by Regione Piemonte - Laboratorio Agrochimico - Settore Fitosanitario e Servizi Tecnico-Scientifici.