Study area and sampling sites
Stopića Cave is located on the left bank of the river Prištavica, at an altitude of 711 m, on the eastern edge of the Zlatibor mountain in western parts of Serbia. The entrance to the cave is 30-40 m wide and is located at the bottom of a 50 m high vertical limestone cliff [19]. Stopića cave consists of five morphological units: Light Hall, Dark Hall, Hall with Tubs, Channel with Tubs and River Channel [19]. Light Hall – the very beginning of the cave (the entrance zone), Dark Hall (transitional zone) and Hall with Tubs are accessible and adapted for tourists. The most famous tourist attraction is the Hall with Tufa Tubs (Rimstone Dams), where many of Tufa Tubs are recognized as the biggest and the deepest (over 7 meters) of all other caves in Serbia [20]. The cave has been protected by the Serbian government as a natural good of the first category since 2005. Sampling was carried out in three seasons during 2020: Winter (mid January 2020, prior to the Covid-19 pandemic lockdown), Spring (early June, at the beginning of the tourist season peak, after Covid-19 pandemic lockdown), Summer (early September, at the end of the tourist season peak) and in four selected sites within the cave: Entrance, Crossroads, Tufa Baths and Waterfall (Figure 1).
Measurement of microclimate parameters
Environmental parameters, temperature (T, °C) and relative humidity (RH, %), were measured in situ using temperature/humidity meter (VELLEMAN DEM105).
Air sampling and estimation of propagule concentrations
Airborne fungi were sampled by the volumetric air sampling method using the SAS Super DUO 360 Air Sampler. Cave indoor air (100 L) was vacuumed and subsequently inoculated on three different mycological media: Potato Dextrose Agar (PDA) and Sabouraud dextrose agar (SDA) for mesophilic fungi and Malt Yeast 40% sucrose agar (M40Y) for xerophiles and xerotolerants. Before each sampling, the sampler was sterilized with ethanol (70 %) to prevent cross-contamination.
Petri dishes with inoculated nutrient media were transported to the laboratory and incubated in a thermostat (UE 500, Memmert) at 25 ± 2 °C for a week. After the incubation period, the Petri dishes were examined, and all visible colonies were counted and labeled. The obtained numbers of the colonies were then converted to probable statistical total values, and calculated by formula given by Feller [21] and expressed as CFU m-3 of air. All morphologically different colonies were inoculated in order to obtain pure cultures.
Identification of fungi
Isolated fungi were inspected using stereomicroscope (Stemi DV4, Carl Zeiss) and light microscope Zeiss Axio Imager M.1) and identified according to morphological criteria (colony morphology and microscopic characteristics of reproductive structures) using identification keys: Watanabe [22] and Samson et al. [23]. Molecular methods are used to confirm preliminary identification, as well to identify non-sporulating isolates. In that sense primary fungal isolates were reinoculated on PDA, SDA and M40Y media and incubated at 25 ± 2 °C for a week. For DNA extraction dry peripheral mycelia (approx. 40 mg) were harvested according to the manufacturer’s instructions of a DNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA). PCR amplification of the ITS region and BenA gene were carried out using selected ITS1/ITS4 [24], and Bt2a/Bt2b primers [25], respectively. For ITS region PCR amplification was accomplished through: initial denaturation (95 ◦C; 4 min), 35 amplification cycles (95 °C; 30 s / 52 °C; 1 min / 72 °C; 1 min) and final extension (72 °C; 7 min). On the other hand, PCR amplification of BenA was conducted as follows: initial denaturation (95 °C; 4 min), 35 amplification cycles (94 °C; 1 min / 58 °C; 1 min / 72 °C; 1 min) and final extension (72 °C; 7 min). Amplification reactions were conducted in the SuperCycler Thermal Cycler SC300T (Kyratec, Brisbane, Australia). The reaction mixture (50 μl) contained 2 μL DNA, 25 μL 2X DreamTaqTM Green PCR Master Mix (Thermo Fisher Scientific, Vilnius, Lithuania) and 19 μL water, nuclease-free (Thermo Fisher Scientific, Vilnius, Lithuania). The amplified DNA fragments were fractionated in agarose gels (1%) in 0.5 × TBE buffer. Midori Green stain was used for DNA visualization by UV illumination [26]. The obtained PCR products were then shipped for purification and sequencing to Macrogene (Netherlands). The resulting sequences were then compared with other related sequences deposited to the National Center for Biotechnology Information (NCBI) using the BLAST program (BLAST+ 2.7.1 of the NCBI). Finally, obtained fungal DNA sequences were then deposited in the relevant GenBank database of NCBI.
Phylogenetic analysis of airborne fungal communities
Sequence alignment was carried out using the CLUSTALW algorithm in MEGAX software [27]. The phylogenetic tree was built on the basis of the alignment and DNA sequences comparison by employing Maximum likelihood phylogeny (1000 bootstrap replicas). Kimura 2 parameter model was determined as the best for estimating genetic distances between tested sequences – measured in the terms of nucleotide substitutions per site. Rhizophydium brooksianum JEL 136 (NR_119550.1) was used as the outgroup.
Ecological indicators
In order to estimate indoor air quality in the “Stopića pećina” cave, obtained results of the concentration of fungal propagules in the different cave rooms were expressed in CFU m-3 and compared with ecological indicators proposed by Porca et al. [1]. According to these authors, the true cave atmosphere has been classified into five categories based on fungal propagules concentration: 1. Caves with fungal concentration less than 50 CFU m-3 (Category I – indicating no problem with fungi); 2. Caves with fungal concentration between 50 and 150 CFU m-3 (Category II – requires some periodic controls and studies to eliminate fungal problem); 3. Caves with fungal concentration between 150 and 500 CFU m-3(Category III – threatened by fungi and requires different cave management and controls); 4. Caves with fungal concentration between 500 and 1000 CFU m-3 (Category IV – already affected by fungi as a result of massive visits or spillage); and 5. Caves with fungal concentration above 1000 CFU m-3 (Category V – have irreversible ecological disturbance).
Fungal communities by throphic mode and ecological guild
All fungal isolates were sorted using FUNGuild v1.0 tool [28] (Guilds_ v1.1.py script, database: http://stbates.org/funguild_db.php) and literature data [23, 29- 33]. Species were classified on the basis of their throphic mode into ecological categories: Pathotroph (P), Saprotroph (S) and Symbiotroph (Sy), and to corresponding ecological guilds: animal pathogen (ap), endophyte (en), epiphyte (ep), fungal parasite (fp), lichen parasite (lp), litter saprotroph (ls), soil saprotroph (ss) plant pathogen (pp), undefined saprotroph (us) and wood saprotroph (ws).
Statistical analyses
Statistical analyses were performed using software Canoco for Windows [34], Microsoft Excel and the statistical package XLSTAT [35].
One-way ANOVA was done to see potential effect of seasonality, sampling position (section of the cave) and growth media on CFU m-3.
Canonical correspondence analysis (CCA) was done to see the relationship between documented fungal genera (presence/absence) and sampling season that is used as explanatory variable, while temperature, relative air humidity and CFU m-3 were used as supplementary variables. Documented fungal taxa were observed also against variables referring to cultivation medium and part of the cave where sampling was done, but those analyses showed no significance.
Categories (CI-CV) of cave atmosphere based on fungal propagules concentration proposed by Porca et al. [1] were also observed in relation to season, part of the cave and medium using constrained analysis (CCA). Only analysis that show categories (CI-CV) of cave atmosphere in relation to sampling season was significant.