Unravelling the impact of light spectra on microalgal growth and biochemical composition using principal component analysis and artificial neural network models

• Published in: Algal research (Amsterdam) 2025-01, Vol.85, p.103820, Article 103820
• Main Authors: Esteves, Ana F., Pardilhó, Sara, Gonçalves, Ana L., Vilar, Vítor J.P., Pires, José C.M.
• Format: Article
• Language: English
• Subjects: Artificial neural network; Biomass composition; biomass production; bioremediation; blue light; carbohydrates; carotenoids; Chlorella vulgaris; chlorophyll; Compound accumulation; exposure duration; Genetic algorithm; Light wavelengths; lipids; Microalgae; neural networks; nutrient uptake; Photosynthesis; principal component analysis; protein content; protein synthesis; red light; wavelengths; white light
• ISSN: 2211-9264; 2211-9264
• DOI: 10.1016/j.algal.2024.103820

Optimising cultivation conditions is essential for making the large-scale production of microalgae-based products economically and environmentally viable. However, the species-specific responses of microalgae to light spectra remain underexplored, particularly regarding the interconnected effects on growth, nutrient uptake, photosynthetic performance, and biochemical composition. This study presents the impact of different light spectra on Chlorella vulgaris growth and nutrient uptake. The photosynthetic activity and biomass biochemical composition were also assessed at two different growth stages (late-exponential and stationary phases). Principal component analysis (PCA) was also conducted to uncover the relationships between key variables, including light wavelength, exposure time, biomass concentration, nutrients availability, N:P ratio and biomass composition. Additionally, genetic algorithm-optimised artificial neural networks (GA-ANN) models were developed to predict biochemical composition based on environmental variables. The results demonstrated that red light promoted high growth rates (0.509 ± 0.004 d−1) and biomass productivity (103 ± 6 mg L−1 d−1), whereas blue light worsened growth-related results (0.424 ± 0.003 d−1). NO3-N uptake was enhanced by white light (21.6 ± 0.4 %) and orange light in the stationary phase (32 ± 1 %). Meanwhile, PO4-P uptake was boosted by red light in the late-exponential phase (67 ± 2 %). Regarding the biomass composition, blue light enhanced protein production (33.8–39.0 % DCW), whereas red light increased carbohydrate accumulation (20.2–23.5 % DCW). The lipid (21.6–24.5 % DCW) and photosynthetic pigment contents were boosted by white light. The GA-ANN models demonstrated strong predictive accuracy, with protein content showing the highest performance (R2: 0.991, RMSE: 0.006 % DCW), followed by carotenoids and chlorophyll content. The outcomes of this study are useful for improving microalgal production techniques, bioremediation strategies and target compound accumulation methods. [Display omitted] •Light spectra should be selected based on the goal of microalgal production.•Red light boosted microalgal growth, nutrient uptake and carbohydrate accumulation.•Blue light increased protein content.•Orange light boosted NO3-N and PO4-P uptake.•White light enhanced NO3-N uptake, lipid and photosynthetic pigment production.