Embedding response surfaces in kinetic models predicts microalgae growth across light and nutrient gradients

Area of Science:

• Biotechnology
• Algal Biotechnology
• Bioprocess Engineering

Background:

• Microalgae are key for sustainable biofuel, high-value compounds, and wastewater treatment.
• Accurate prediction of microalgal biomass productivity is crucial for industrial applications but challenging due to varying light and nutrient conditions.
• Existing models have limitations: mechanistic models assume constant parameters, while response surface methodology (RSM) lacks dynamics and can produce implausible results.

Purpose of the Study:

• To develop a hybrid model integrating biologically interpretable Monod-type surfaces into a logistic growth model.
• To enable maximum growth rate and carrying capacity to vary continuously with light intensity and nutrient availability.
• To provide a robust tool for predicting microalgal productivity and optimizing cultivation.

Main Methods:

• Developed a hybrid model combining logistic growth with Monod-type surfaces for dynamic parameter variation.
• Validated the model using 1315 growth curves of Chlamydomonas reinhardtii under mixotrophic conditions.
• Assessed model performance at flask (50mL) and microplate (200µL) scales, comparing it against established mechanistic models.

Main Results:

• The hybrid model accurately predicted time-series growth under novel light-nutrient conditions at flask (R²=0.94) and microplate (R²=0.84) scales.
• It outperformed three established mechanistic models in predictive accuracy.
• The model reduced the number of parameters from up to 50 to six global parameters, decreasing experimental effort.
• Microplate-derived parameter surfaces were successfully mapped to flask-scale surfaces (R²≥0.94), demonstrating scale-up potential.

Conclusions:

• The hybrid model offers a practical and accurate framework for optimizing microalgal cultivation at the laboratory scale.
• It provides a significant reduction in experimental requirements compared to traditional methods.
• The model's transferability across different culture volumes suggests its utility for scaling up microalgal production.