Quantifying cyanobacterial phycocyanin concentration in turbid productive waters: A quasi-analytical approach

In this research, we present a novel technique to monitor cyanobacterial bloom using remote sensing measurements. We have used a multi-band quasi analytical algorithm that determines phytoplankton absorption coefficients, aϕ(λ), from above surface remote sensing reflectance, Rrs(λ). In situ data including remote sensing reflectance, phytoplankton pigment concentration, and absorption coefficients of optically active constituents in the water were collected from highly turbid and productive aquaculture ponds. These shallow (< 1.5 m) ponds in northwestern Mississippi, USA, were used for channel catfish Ictalurus punctatus aquaculture and had high nitrogen and phosphorus loading rates from manufactured feeds added to ponds to promote rapid fish growth. These practices resulted in high phytoplankton biomass (chlorophyll-a concentrations = 59.4–1376.6 mg m− 3) with communities dominated by filamentous, gas-vacuolate cyanobacteria. A novel technique was developed to further decompose the aϕ to obtain phycocyanin absorption coefficient, aPC, at 620 nm, a primary peak of phycocyanin absorption spectrum. Validation of the model produced mean and median absolute relative errors of 36.2% and 22.0%. Overall, the model performance was higher in the higher range of PC concentration (> 150 μg l− 1). Results demonstrate that the new approach will be suitable for quantifying phycocyanin concentration in cyanobacteria dominated turbid productive waters.

► We propose a novel technique for remote quantification of cyanobacterial phycocyanin.
► We have developed an analytical scheme to correct for the effects of chlorophyll-a.
► Mean and median absolute relative errors of the new model were 36.2% and 22.0%.
► Accuracy of the new model was considerably higher than the best existing algorithm.