Source resolution and model prediction of hydrogen peroxide produced through photochemistry and phytoplankton species...

Hydrogen peroxide (H₂O₂) plays a crucial role in maintaining a chemical and ecological equilibrium in aquatic environments. However, there is still a lack of comprehensive understanding regarding the potential drivers of H₂O₂ production. This study investigates the source and potential drivers of H₂O₂ production in natural aquatic systems, revealing its concentrations as a synergistic outcome of photochemical and biological processes, and establishes a multiparameter predictive framework. By analyzing dissolved organic matter (DOM) extracted from four lakes (Fan Shen, Cai Yun, Jiu Long, Bao Sheng) through water quality parameters, spectral characteristics, and molecular composition, combined with artificial solar irradiation experiments, we demonstrate that DOM abundance and low-molecular-weight components promoted photochemical H₂O₂ generation via photosensitized reactions, while total phosphorus (TP) suppresses photochemical production through organophosphorus-H₂O₂ interactions. Biological assays with Microcystis aeruginosa, non-toxic Microcystis spp, and Chlorella vulgaris reveal that proteins-/lipids-rich DOM significantly elevates biogenic H₂O₂ generation by phytoplankton species. Chlorella vulgaris which exhibited the highest ability to generate H₂O₂ was the most resilient to oxidative stress caused by H₂O₂. Combined photochemical-biogenic (P&B) experiments confirmed that synergistic H₂O₂ concentrations (combined P&B H₂O₂) were lower than the biogenic H₂O₂ concentrations due to algae experiencing oxidative stress from photochemical H₂O₂ and activating its decomposition capacity, with the Variance Partitioning Analysis (VPA) identifying that H₂O₂ production was influence by a combination of water quality, DOM spectral traits and biological factor (34.45-55.45 %), which was higher than the influence of individual factors (3.21-7.07 %). A multiple multiparameter regression model incorporating TOC, E2/E3, TP, and microbial-derived tryptophan (C2) achieves robust H₂O₂ prediction across eight lakes. By elucidating the DOM-algae mechanistic interplay governing H₂O₂ dynamics and modeling H₂O₂ concentrations, our study establishes an operationalized framework for assessing oxidative stress and predicting algal blooms, providing critical tools to the early warning of algal bloom events in aquatic ecosystems.