A fluorometric differentiation technique of phytoplankton assemblage based on Coif2 wavelet and fourth-derivative

Abstract

The 3D discrete fluorescence spectra with 12 excitation wavelengths (400, 430, 450, 460, 470, 490, 500, 510, 525, 550, 570, 590nm) were determined by fluorescence spectrophotometer for 20 phytoplankton species. Then, the wavelet (Coif2), fourth-derivative and non-negative least squares were applied to establish the fluorescence differentiation method which could differentiate phytoplankton populations at the levels of both division and genus. This method was tested: for simulative mixed samples (the dominant division algae accounted for 50%, 70%, 90% and 100% of the gross biomass, respectively) at the level of division, the discrimination rates were 94.4%, 97.8%, 98.6%, and 98.4% with average relative content of 48.2%, 65.1%, 79.4% and 77.9%, respectively. For simulative mixed samples(the dominant species accounted for 70%, 80%, 90% and 100% of the gross biomass, respectively) at the level of genus, the correct discrimination rates were 72.7%, 82.2%, 86.5% and 81.0%, respectively. For the in situ test, all of the 12 samples were recognized at the division level, and for the three samples which the dominant species accounted for more than 80% of the gross biomass, the dominant species of one was recognized at the genus level. As a result, an in situ algae fluorescence auto-analyzer which uses a series of LEDs as the light sources is developing. The technique also can be directly applied on fluorescence spectrophotometer.