Analysis of Phytoplankton Pigments by Excitation Spectra of Fluorescence

The advantages and problems of the application of actively excited fluorescence in natural phytoplankton analysis are discussed. The focus is made to a correct prediction of pigment concentrations by fluorescence data. The results of high resolution mapping of chlorophylls and phycobilins in the Gotland Basin (Baltic Sea) during Cyanobacterial blooms in 1997 and 1998 are presented. Dynamical spatial maps of phytoplankton pigments were registered on line with the flow-through spectrofluorimeter FLUO-IMAGER and the shipboard laser remote sensing spectrometer FLS-S. Characteristic patterns of the phytoplankton distribution in the study area and their evolution in time are discussed. Field studies confirmed that on line spectrofluorimetry can be effectively used to diagnose phytoplankton pigments in application to the tasks of phytoplankton bloom detection, analysis of phytoplankton development and spatial mapping of photosynthetic pigments. Considering the tasks of monitoring large water areas or temporary processes in a spot area, the most productive way is a balanced combination of on line continuous fluorescence measurements and sampling procedures, which allows to decrease the time-consuming manual analysis of water samples in the laboratory. INTRODUCTION Fluorescence methods are widely applied for characterization of the phytoplankton community status in vivo in the marine and freshwater environment. Various techniques are used to estimate the concentration of chlorophyll a, biomass and primary production. Based on the method to induce the fluorescence of photosynthetic pigments of phytoplankton in vivo in a discrete sample or directly in a water habitat different scientific systems and commercial instruments exist. The broadest class of devices that began to be applied in phytoplankton analysis was connected with measurements of fluorescence based on fixed wavelength fluorimetry with a filter system for excitation and detection of fluorescence (1), and scanning spectrofluorimetry based on recording the excitation or emission spectra of fluorescence (2). In the last decade a number of new fluorescence techniques has been intensively developed, including fast repetition rate fluorometry (3) and the pump and probe technique (4). The common feature of all these techniques is that the concentration of pigments is estimated through the correlation of in vivo fluorescence with the isolated photosynthetic pigment values. Attempts to quantify the phytoplankton pigments in situ using fluorescence characteristics run against the fact that photosynthetic pigments in a living cell are bound to protein. Due to that, the in vivo fluorescence of phytoplankton should be considered in connection with the specific pigment-protein structure of the cells. Various pigment-protein units in a cell have a different structure and play a different role in photosynthesis (5). Some of the pigment-protein units are located in photosynthetic reaction centres and participate in the light reactions of photosystem I and photosystem II. The bulk of the chlorophyll/carotenoid-protein consists of various light-harvesting pigment-proteins whose role is to collect light and transfer the absorbed energy to the reaction centres. Therefore, the characteristics of phytoplankton in-vivo fluorescence, including its fluorescence efficiency, depend on the taxonomy. Proceedings of EARSeL-SIG-Workshop LIDAR, Dresden/FRG, June 16 – 17, 2000 EARSeL eProceedings No. 1 225 There have been numerous attempts to achieve a taxonomic identification based on in vivo biooptical characteristics of phytoplankton, including fluorescence excitation and absorption spectra. Earlier investigations, introducing the chlorophyll a/accessory pigment ratio technique (1,6) in excitation spectra, have shown that the identification of the alga class or species in natural mixed phytoplankton population is quite problematic. Fluorescence becomes a powerful tool in cell identification when the cells can be separated and analysed individually, e.g. with flow cytometry (7,8). Phytoplankton in vivo fluorescence excitation spectra depend not only on the taxonomic position of algae, but also on the photoadaptation state. The cellular pigment content, the ratio of total chlorophyll a to accessory pigments, and the efficiency of energy transfer to chlorophyll a are sensitive to the light condition of culture growth (9,10,11). The effect of photoadaptation on fluorescence excitation spectra of algae can be of the same level of significance as the influence of taxonomy (12). In particular, photoadaptation effects must be taken into account in interpreting vertical profiles of phytoplankton, when the gradient of ambient light may be quite high. The fluorescence yield depends also on the stage of growth of phytoplankton (13). The analysis of fluorescence of different alga species at different stages of growth in controlled light and nutrient conditions has shown that the predicted chlorophyll by fluorescence can be lower than the real concentration at the growth stage of phytoplankton. This is mainly due to a change of the cellular pigment content. As the intensive growth of phytoplankton is typically registered in bloom processes, it is important to quantify such development when considering the task of diagnostics of phytoplankton bloom (14). In spite of the complicated correlation of in-vivo fluorescence and phytoplankton pigments further development of the technique continued to distinguish the fluorescence of different pigments in the phytoplankton community (15). Species/class-specific bio-optical and photoacclimation characteristics were overviewed and related to the pigment composition of phytoplankton for major marine phytoplankton groups (16). These numerous studies were connected with the definition of classspecific marker pigments for the indication of bloom-forming and toxic species. Due to the fact that phycobiliproteins of cyanobacteria are strongly auto-fluorescent and do not induce a dominant chlorophyll a signal, the fluorescence technique was successfully used for vertical and horizontal in-situ profiling of phycobiliproteins in the marine environment (17). Up to now the method of in vivo fluorescence excitation spectra remains the most specific to the variability of pigment composition among other bio-optical techniques to characterise living phytoplankton and could be used for on line monitoring (18). The excitation spectra are little affected by light scattering (19) and do not depend on absorption of light by photoprotective pigments (20). Recent laboratory investigations have shown that the ratio of spectral peaks in excitation spectra of fluorescence are less sensitive to growth stage and light conditions than the fluorescence intensity (21). The aim of the present work was to study the capabilities of the fluorescence excitation spectra technique for a rapid characterisation of living phytoplankton: 1) to reveal the common features of class-specific excitation spectra, 2) to investigate the possibilities of in vivo phytoplankton pigments quantification, 3) to consider the approach of phytoplankton excitation spectra application for online monitoring. MATERIALS AND METHODS Algae collection The culture collection of the Institute of Biology of Southern Seas (Sevastopol, Ukraine) was used. Algae cultures of 32 phytoplankton species covering major marine taxa were studied (Table I). Batch cultures were maintained in a modified Goldberg medium (22) and an enriched seawater medium (23) at room temperature (18-22o C) under daylight. Synecococcus sp. strains were obtained from Black Sea picoplankton. All trials were conducted in the late log-phase or in the stationary phase of the cultures. Proceedings of EARSeL-SIG-Workshop LIDAR, Dresden/FRG, June 16 – 17, 2000 EARSeL eProceedings No. 1 226 Instruments Fluorescence excitation spectra of culture samples were measured with a Fluo-Imager M32B spectrofluorimeter at three emission wavelengths. The first spectral range (excitation 400 to 660 nm, emission 680 nm) was aimed to reveal fluorescence of chlorophyll a induced via accessory pigments. The second spectral range (excitation 400 to 620 nm, emission 630 nm) was used to record separately the phycocyanin (PC) fluorescence. The third spectral range (excitation 400 to 550 nm, emission 580 nm) was used to record the phycoerythrin (PE) fluorescence. A xenon lamp (150 W) was used as a light source. The fluorescence of the sample in a flow-through quartz cuvette is induced via the excitation monochromator and recorded with a photomultiplier tube (PMT) with further digital processing. An interference filter system that contains three filters (678, 630, 580 nm) was used to select the spectral bands for registration. Specially designed software provided spectral data analysis and control of instruments. The excitation spectra were not corrected for the spectral distribution of the lamp source. In the field studies the fluorescence excitation spectra were measured continuously in a flowthrough mode. Water from 3 to 5 m depth was pumped and analysed every 2 minutes providing an averaged spatial resolution of approx. 250 m at a cruise speed of about 7 knots. The software analysed the fluorescence excitation spectra in real-time to reveal the fluorescence intensities corresponding to different phytoplankton pigments. The intensity of pigment fluorescence was stored according to the time and coordinates of the measurements. Table 1. List of phytoplankton cultures. Pigment concentration Seawater samples were filtered through Whatman GF/F glass fiber filters. After breaking the cells with ultrasonication, the pigments were extracted in 96% ethanol. Chlorophyll a was measured using a Shimadzu UV-2101PC spectrophotometer and calculated for concentration according to (24). RESULTS AND DISCUSSION In vivo fluorescence excitation spectra of phytoplankton cultures were measured at the emission wavelength 680 nm. The spectra are shown in Figure 1. Class No Species Label Class No Species Label Bacillariophyceae 1 Thalassiosira pseudonana pl59 Chlorophyceae Stichococcus bacillaris pl63 2 Phaeodactylum tricornutum pl64 3 Chlorella minutissima pl215 4 Chlorella elipsoidea pl216 Dynophyceae 3 Exuviaella pusilla pl61 6 Chlarococcum infus. pl218 4 Prorocentrum cordata pl56 8 Dunaliella salina pl221 5 Gymnodinium kowalewskii pl57 9 Dunaliella viridis pl224 6 Amphidinium klebsii pl55 11 Dunaliella martima pl226 7 Gymnodinium lanskaja pl202 12 Dunaliella tertiol. pl227 9 Peridinium trochoidea pl212 10 Peridinium triquetrum pl213 Prasinophyceae 10 Tetraselmis sp. pl225 13 Platimonas viridis pl223 Chrysophyceae 12 Olistodiscus luteus pl60 Cyanophyceae Coccochloris sp. (LL) pl203 Prymnesiophyceae 13 Pavlova luteri pl65 Coccochloris sp. (HL) pl204 14 Isochrysis galbana pl220 Synechococcus aeruginosus pl205 15 Monochrysis luteri pl228 Oscillatoria sp. pl206 16 Gephyrocapsa huxleyi pl52 Synechococcus sp.1 pl207 17 Emiliania huxleyi pl209 Synechococcus sp.2 pl208 Proceedings of EARSeL-SIG-Workshop LIDAR, Dresden/FRG, June 16 – 17, 2000 EARSeL eProceedings No. 1 227 Figure 1: Fluorescence excitation spectra of phytoplankton cultures at the emission 680 nm. The fluorescence excitation spectra of the cultures can be divided into three types of characteristic spectra according to pigment representation. The blue–green algae belong to the first type that is characterised by the absence of chlorophyll and carotenoid peaks in excitation spectra. The biliproteins play the role of accessory pigments to collect and transfer the light energy to the reaction centre. Biliproteins in blue-green algae are divided into three classes based on the position of their absorption bands: PE, PC, and allophycocyanins (in increasing order of wavelength). PE is displayed in the excitation spectra (Figure 1) as a broad band with a maximum near 560 nm (25), PC – as a shoulder of a band with a maximum near approx. 635 nm (26). In fact, the fluorescence recorded at the wavelength of chlorophyll a emission is a part of biliprotein autofluorescence. The second type of excitation spectra is specific for chlorophyll b–contaning algae that are represented by species of Chlorophyceae. The single broad band in excitation spectra in the range 420-440 nm is connected with chlorophyll a. The spectral band in the range 460-490 nm is caused by a light-harvesting complex containing chlorophyll b and carotenoids. Chlorophyll b absorbs in vivo at 470 nm (27). A sharp decrease of intensity at 500 nm was observed in in-vivo excitation spectra for all Chlorophyceae. The last group comprises chlorophyll c-containing algae. Numerous peaks in different combinations fill the wide spectral range from 400 to 600 nm. Similar to chlorophyll b–contaning algae, the spectral range 420-440 nm is connected mainly with chlorophyll a. The peaks in the range 460470 nm are caused by diverse combinations of chlorophyll c1, c2 and c3. The range 480-580 nm Chlorophyceae