Community-level physiological profiling of carbon substrate metabolization by microbial communities associated to sediments and water in karstic caves from Romania
Cave ecosystems comprise specialized microbial communities that play essential roles in biogeochemical cycles; yet their metabolic capabilities and ecological functions are not yet fully understood. As conventional cultivation techniques offer limited insights into the metabolic capabilities, methods based on direct functionality screening could provide more in-depth knowledge of cave microbial activity. In this study, we utilized the Community-Level Physiological Profiling (CLPP) based on Biolog(R) EcoPlate approach to assessing carbon substrate utilization by microbial communities associated with pool water, sediment, sediment (limon), and moonmilk from five caves in Romania. Principal Component Analysis (PCA) and Generalized Additive Models (GAMs) statistics were employed to infer the patterns of C-substrate metabolization and their environmental drivers. Environmental variables such as sodium (Na) and electrical conductivity (EC) significantly impacted C-utilization capabilities as indicated by both PCA and GAM. The latter analysis elucidated non-linear relationships between variables, such as EC, Na, and Mg, and microbial metabolic diversity indices. However, distinct C utilization patterns were detected among sampled sites and chemical types. Unlike moonmilk samples whose associated microbial communities appeared as exhibiting low C-substrate utilization, the highest activity was shown in cave pool water samples with the associated microbial communities extensively consuming D-galacturonic acid and Tween 80. Conversely, substrates like L-threonine and -ketobutyric acid showed limited utilization across all cave samples. Average Well Color Development (AWCD) and Shannon diversity indices indicated that microbial communities associated to samples from Closani and Muierilor caves demonstrated the highest metabolic diversity. Our findings suggested that metabolic profiling using Biolog(R)EcoPlates method combined with multivariate statistical methods might prove as suitable approach to effectively screen for cave microbial functionality and the probable environmental drivers. Besides, this work distinguishes from similar studies by relying on GAM analysis to predict the environmental factors governing the microbially-mediated organic carbon degradation in subterranean ecosystems.