Dynamic microbial community assembly and molecular ecological network responses in biological activated carbon filters: effects of algal organic matter exposure and empty bed contact time.

Biological activated carbon (BAC) filtration plays a crucial role in advanced drinking water treatment. Recent researches have shifted from decontamination performance evaluation to process optimization and customization of microbial communities. The responses of microbial communities to seasonal water quality variations caused by algal outbreaks or deaths, and operational conditions of filtration medium and empty bed contact time (EBCT), along with dynamics of assembly processes and molecular ecological networks remain insufficiently understood. Herein, the decontamination performance of four BAC columns packed with varied physicochemical properties of granular activated carbon (GAC), exposed to algal organic matter (AOM) and changes of EBCT was investigated. Microbial diversity, assembly mechanisms, and dynamics of molecular ecological networks were systematically evaluated. Results showed that coal-based BAC exhibited superior decontamination performance under AOM exposure, with average removals of COD (47.23 %), UV (55.82 %), and NH-N (65.01 %), along with higher microbial diversity and richness than that of wood-based BAC. AOM exposure increased microbial diversity, while shortened EBCT reduced it. Deterministic processes in community assembly intensified under both AOM exposure and a shortened EBCT of 10 min, the proportion were up to 82 % and 75 %, respectively. AOM exposure increased network scale and complexity, whereas the opposite trend was observed with a shortened EBCT of 10min. Structural equation modeling identified that influent water quality (path coefficient = 1.00) was the dominant driver of microbial diversity, followed by GAC properties (0.30) and EBCT (-0.35). These findings provide insights for microbial community customization and BAC process optimization to control algal-derived organic contamination.