Machine learning-assisted quorum sensing molecules regulate growth progression and bioactivity of an oxygenic photobiofilm for non-aeration greywater purification.

Microalgal-bacteria consortium provide self-aeration through photosynthetic oxygen production, offering an energy-efficient strategy for wastewater treatment. However, biofilm regulation in practical application is still limited by various factors, with acyl-homoserine lactones (AHLs) serving as key regulatory elements. This study established a microalgal-bacteria biofilm reactor (MBBfR) to investigate the regulatory mechanisms of AHLs-mediated quorum sensing under varying nitrogen (N) source. Results showed that a machine learning (ML) model successfully predicted the total suspended solids (TSS, R2 = 0.951), specific oxygen consumption rate (SOCR, R2 = 0.873) and oxygen generation rate (SOGR, R2 = 0.977) using extracellular polymeric substances (EPS) and AHLs as predictors. N source and C/N ratio altered AHLs concentrations and their associations with microbial community composition, with C6-HSL identified as the predominant AHLs. Nitrate facilitated the rapid formation of high quality and density biofilms, and enhanced signaling molecules secretion, thereby improving N removal efficiency. With high N stress and low C/N ratio, AHLs significantly induced EPS production, maintained microbial metabolic and photosynthetic activities, and promoted microalgal viability and proliferation. This signaling mechanism contributed to sustaining MBBfR stability and function. The findings provide a theoretical foundation and practical guidance for optimizing MBBfR-based wastewater treatment, enabling precise and convenient biofilm functionality regulation.