Abstract
The effective identification of bacterial and fungal isolates is essential for microbiological monitoring in environments like speleotherapeutic caves. This study compares MALDI-TOF MS and the OmniLog ID System, two high-throughput culture-based identification methods. MALDI-TOF MS identified 80.0% of bacterial isolates to the species level, while the OmniLog ID System identified 92.9%. However, species-level matches between the methods were only 48.8%, revealing considerable discrepancies. For discrepant results, MALDI-TOF MS matched molecular identification at the genus level in 90.5% of cases, while the OmniLog ID System matched only in 28.6%, demonstrating MALDI-TOF MS's superiority. The OmniLog ID System had difficulties identifying genera from the order Micrococcales. Fungal identification success with MALDI-TOF MS was 30.6% at the species level, potentially improvable with a customised spectral library, compared to the OmniLog ID System's 16.7%. Metagenomic approaches detected around 100 times more microbial taxa than culture-based methods, highlighting human-associated microorganisms, especially Staphylococcus spp. In addition to Staphylococcus spp. and Micrococcus spp. as indicators of cave anthropisation, metagenomics revealed another indicator, Cutibacterium acnes. This study advocates a multi-method approach combining MALDI-TOF MS, the OmniLog ID System, culture-based, and metagenomic analyses for comprehensive microbial identification. Metagenomic sampling on nitrocellulose filters provided superior read quality and microbial representation over liquid sampling, making it preferable for cave air sample collection.