Copper stress responses in Scenedesmus obliquus-Bacillus subtilis Consortia: Machine Learning-Based prediction of...

Copper pollution in wastewater is an environmental issue that requires efficient and sustainable waste treatment methods. To improve the efficiency of traditional microalgal treatment of heavy metals and enable predictive assessment of treatment outcomes, this study examined the physiological behavior of Scenedesmus obliquus to copper stress and the mitigating role of a microalgae-bacteria consortium. In addition, the study assesses the usefulness of physiological and biochemical responses as inputs for machine learning models to predict copper removal efficiency under varying stress conditions. The findings showed that copper (Cu) exposure inhibited microalgae growth, photosynthetic performance, and antioxidant defense, while increasing the oxidative stress marker malondialdehyde (MDA) accumulation in S. obliquus. Cu removal efficiency was comparable at 20 ppm, with 83.4% obtained in the microalgae group and 85.2% in the consortium group. It was notably higher at the highest Cu level of 80 ppm in the consortium group (75.4%) than in the algal group (69.0%). The machine learning model trained on physiological and physicochemical inputs achieved excellent predictive performance (R² = 0.9744), with deviations within ± 3%. These findings highlight the S. obliquus-B. subtilis consortium as a resilient system that combines improved tolerance and copper remediation efficiency with the predictive power of machine learning to offer a potential advanced bioremediation strategy.