Synergistic biochar and Serratia marcescens tackle toxic metal contamination: A multifaceted machine learning approach

Metal contamination in soil poses environmental and health risks requiring effective remediation strategies. This study introduces an innovative approach of synergistically employing biochar and bacterial inoculum of Serratia marcescens to address toxic metal (TM) contamination. Physicochemical, enzymatic, and microbial analyses were conducted, employing integrated biomarker response (IBR) and machine-learning approaches for toxicity estimation. The combined application significantly reduced the Cd, Cr, and Pb concentrations by 71.6, 31.2, and 57.1%, respectively, while the Cu concentration increased by 85% in the individual Serratia marcescens treatment. Biochar enhanced microbial biomass by 33-44% after 25 days. Noteworthy physicochemical improvements included a 44.7% increase in organic content and a decrease in pH and electrical conductivity. The K⁺ and Ca²⁺ concentrations increased by 196.9 and 21.6%, respectively, while the Mg²⁺ content decreased by 86.4%. Network analysis revealed intricate relationships, displaying direct and indirect negative correlations between metals and soil physicochemical parameters. The IBR index values indicated effective mitigation of TM toxicity in Serratia marcescens and biochar with individual and combined treatments. Binary classification demonstrated high sensitivity (80.1%) and specificity (80.5%) in identifying TM-contaminated soil. These findings indicate significant biochar- and Serratia marcescens-induced impacts on toxic metal availability, physicochemical properties, and enzymatic activities in metal-contaminated soil, suggesting that blending soil with biochar and microorganisms is an effective remediation strategy.