A Spectral Model of Phytoplankton Absorption for Inland Lake

Abstract

Cyanobacteria (primarily Microcystis aeruginosa; MA) blooms, caused by eutrophic waters, have a significant and negative effect on human health and on the ecosystem as a result of blue-green algal toxins. In this study, a novel phytoplankton absorption spectral model was established using normalized MA and Scenedesmus obliquus (SO) absorption spectra as reference vectors to characterize the absorption properties of cyanobacteria in an inland lake (Dongting Lake, China). This model can be used to estimate the relative contributions of cyanobacteria to the phytoplankton population and thus to phytoplankton absorption. The ability of the model to accurately estimate the percentage of MA in the phytoplankton population and to simulate phytoplankton absorption spectra was examined using in situ data, collected during four campaigns at Dongting Lake. The results show that the absorption characteristics of phycoerythrin and phycocyanin were well simulated by the model in highly turbid and eutrophic inland waters. The percentage of mean square root (RMSP) of the model is 0.71–11.90%, while the model parameters (Sm and 1 − Sm) linearly correlated to the percentage of MA (R2 = 0.70) and SO (R2 = 0.94). Sm is highly correlated with the traditional cyanobacterial percentage indicator and provides a proxy for the relative contribution of cyanobacteria to phytoplankton absorption. This phytoplankton absorption model fully considers the effect of accessory pigments, which are important for the identification and monitoring of cyanobacterial blooms. The results show that the model can successfully retrieve the composition ratio of harmful algae (MA) and well reconstruct the absorption characteristics of phycocyanin.