Quantitative image analysis for assessing extracellular polymeric substances in activated sludge under atrazine exposure

• Main Author: Melo, Antonio
• Publication Date: 2024
• Other Authors: Costa, Joana Sofia Gomes, Quintelas, Cristina Maria Catalão, Amaral, António Luís, Ferreira, Eugénio C., Mesquita, Daniela P.
• Format: Article
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: https://hdl.handle.net/1822/91576
• topic: Atrazine; Sequencing batch reactor; Aggregates; Image analysis; Partial least squares

Extracellular polymeric substances (EPS) play a vital role in biological wastewater treatment systems. This study investigates the impact of herbicide atrazine (ATZ) on the overall performance, EPS yield, composition, and sludge morphology in an activated sludge (AS) system operated in a sequencing batch reactor (SBR). Since conventional methods for analyzing EPS are time-consuming and releases residues, a new approach was developed in this work to evaluate the EPS fractions and components, based on the morphological characterization of the biomass using quantitative image analysis (QIA) technique coupled with multivariate statistics. Results showed that exposure to ATZ inhibit biomass activity in terms of organic matter (COD) and nitrogen removal. Moreover, both tightly bound EPS (TB-EPS) and loosely bound EPS (LB-EPS) increased under ATZ, indicating that microorganisms release EPS as a defense mechanism against environmental changes. The PN/PS ratio also increases, indicating likely increased hydrophobicity in ATZ phases. Furthermore, ATZ phases exhibit a predominance of large aggregates compared to intermediate and small ones, indicating a change in aggregate morphological structure associated with EPS production. The new approach using QIA coupled with partial least squares (PLS) modeling provides accurate predictions of EPS content. The increase in TB-EPS is closely related to the rise of large aggregates in phases exposed to higher ATZ concentrations. The PLS models demonstrate high accuracy for EPS prediction (coefficients of determination, R2 above 0.86), showcasing the feasibility of using QIA for EPS assessment in AS systems. This approach offers significant potential for regular process monitoring and management, providing a more environmentally friendly methodology by eliminating the need for chemical usage.