Integrating PARAFAC and spectroscopic indices to characterise algal organic matter across the growth phases of three bloom-forming algal species toxic to aquaculture

Algal blooms are recognised as a global threat to aquaculture and fisheries, with some blooms being more harmful than others. Hence, it is important for aquaculture facilities to get advanced warning about the occurrence of such blooms. Therefore, in this paper, we developed an integrated framework combining parallel factor analysis (PARAFAC) with spectroscopic indices to identify, characterise and age microalgae species using their organic matter excitation emission matrices (EEMs). We first applied PARAFAC to EEMs obtained from pure cultures of Alexandrium tamarense, Lingulodinium polyedra and Pseudo-nitzschia fraudulenta, with the reference signatures being characterised into protein-like and humic-like components. We observed that the excitation-emission wavelengths corresponding to each component varied, and species differentiation was achieved when tested using known mix samples, with one misclassification noted. Additionally, we applied spectroscopic indices analysis to the pure cultures as a measure of tracking changes in the growth phase and cell density that may influence variation in the intensity of the components. Our regression analysis indicates a strong correlation between the fluorescence indices, biological index and cell densities for two of the species, thus associating the changes in algal organic matter (AOM) production with cell density and growth phase. Similarly, Spearman's correlation revealed a strong relationship between the Yeomin fluorescence index and the protein-like component, indicating that the index value may be associated with AOM protein production. Overall, these results clearly indicate that combining PARAFAC with spectroscopic indices can provide a robust approach for monitoring algal blooms and aid in developing early warning systems.