Algal Interaction-Mediated Biogenic Volatiles Enable Accurate Algal Bloom Prediction

Area of Science:

• Environmental Science
• Biochemistry
• Ecology

Background:

• Algal volatile organic compounds (AVOCs) are key metabolic signals for predicting algal blooms in single-species systems.
• Interspecies interactions complicate algal growth dynamics, necessitating evaluation of AVOC-based prediction under these conditions.

Purpose of the Study:

• To investigate how interspecies interactions between *Microcystis aeruginosa* and *Chlorella vulgaris* influence AVOC profiles.
• To develop machine learning models for predicting algal density using AVOCs under monoculture and coculture conditions.
• To identify specific AVOCs crucial for predicting algal bloom dynamics influenced by interspecies interactions.

Main Methods:

• Proton transfer reaction time-of-flight mass spectrometry (PTR-TOF-MS) to profile AVOCs from monocultures and cocultures.
• Machine learning models, including extreme gradient boosting, to predict algal density.
• Shapley additive explanation and transcriptomic analyses to identify and understand the role of predictive AVOCs.

Main Results:

• Coculture conditions altered AVOC profiles, with compounds linked to fatty acid metabolism and carotenoid degradation observed.
• A machine learning model achieved high accuracy (R²: 0.96) in predicting algal density using AVOCs.
• Interaction-associated AVOCs, particularly terpenoids like DMNT, showed a significant increase in predictive importance (123%) for algal biomass.

Conclusions:

• AVOCs, especially those involved in stress signaling and interspecific communication, are vital for accurate algal bloom prediction amidst species interactions.
• The study provides a framework for real-time bloom prediction by integrating interaction-mediated volatiles into early warning systems.
• DMNT content was validated as a reliable predictor of algal biomass in a natural lake setting.