**3.3. Differentiation of cyanobacterial cultures on the base of single-cell fluorescence spectra**

The automatic on-line differentiation of cyanobacterial species is a key problem in both industrial biomass production and environmental monitoring. In this section, we present a novel technique for taxonomic discrimination of cyanobacteria based on the numerical analysis of *in-vivo* single-cell fluorescence spectra. An optimal set of the parameters is considered, which is sufficient for determination of the taxonomic position of cyanobacteria by means of mathematical statistics. On the base of the linear discriminant analysis obtained spectroscopic data for 21 cyanobacterial strains from *CALU* collection were analyzed. It was shown that the presented technique allows an accurate differentiation of cyanobacteria up to the species/ strain level and enables to distinguish automatically potentially harmful strains.

Since the early 1950s, three different methods are commonly used to characterize phytoplankton and cyanobacterial samples taxonomically: high performance liquid chromatography (HPLC) [13–15, 23, 93]; flow cytometry [10, 11, 94, 95]; and optical microscopy. Various methods have been developed with the aims of increasing accuracy and yielding qualitative information. However, all of them have different limitations. Till now, the best taxonomic differentiation is still obtained using classical inverted microscopy. Unfortunately, this method is time-consuming, human-based and requires appropriate technical skills, and this eliminates the possibility of its application for continuous on-line monitoring. Nearly single-cell flow cytometric analysis is based on light scattering by the cells and fluorescence of the chlorophylls and the phycobilins. It can be easily automated, but it is appropriate only for unicellular species and is useless for numerous industrially cultured filamentous strains. HPLC is the only method, of the three, that is based on the chemical constituents in the sample. The problem is that during the chemical sample preparation, most of the information about the peculiarities of individual species is lost and the residual part of the information is not enough for species/strain classification inside cyanobacterial genera, and is suitable only for the rude differentiation of big classes of phytoplankton. As it was mentioned earlier, several factors contribute to the spectroscopic properties of the phycobilins: the number and chemical nature of the bilins attached to the polypeptide chains; the effects of protein conformation or aggregation state; and interaction between the bilins. Any of this feature can be unpredictably changed during the extraction and purification procedure [96]. Thus, only spectroscopic properties of the intact living cells can give pure unspoiled information about distinctive features of light harvesting complex in specified cyanobacterial strain.

Y(II) decrease in the sonicated culture, which indicates that the physiological state of the culture under sonication is depressed. At the same time, the nonphotochemical quenching of the absorbed light by the fluorescence rises considerably for the treated culture. Comparison of the results of the CLSM spectroscopic measurements with those obtained using conventional fluorimeter and pulse-amplitude modulation approaches confirmed the inhibitory effect of low ultrasonic frequencies (~60 kHz) on the physiological state of cyanobacterial cells and

It can be concluded with confidence that the results obtained via the CLSM technique are correct, and the reduction of the photosynthetic activity and dumping of single-cell physiological state occur as a reply on the external ultrasonic action. The results presented here demonstrate the experiments conducted with the strain Synechocystis CALU 1336 aquatilis; however, similar results were already obtained for another unicellular cyanobacterial species (*Microcystis CALU 398*). Thus, the treatment presented here may refer to a rather diverse

It should be noted that ultrasonic treatment is widely used not only for inhibition of cyanobacteria growth diring harmful blooms, but also for enhancing of protein content and the whole biomass in the industrially cultured strains [91, 92], depending on power-frequency characteristics. Thus, it is very important to obtain on-line correct information about the influence of the ultrasound with given power and frequency on the specified cyanobacterial strain. The noninvasive fluorescent technique presented here gives the opportunity to detect any weak variations in the physiological state of single cyanobacterial cells in real time during sonication.

**3.3. Differentiation of cyanobacterial cultures on the base of single-cell fluorescence** 

strain level and enables to distinguish automatically potentially harmful strains.

The automatic on-line differentiation of cyanobacterial species is a key problem in both industrial biomass production and environmental monitoring. In this section, we present a novel technique for taxonomic discrimination of cyanobacteria based on the numerical analysis of *in-vivo* single-cell fluorescence spectra. An optimal set of the parameters is considered, which is sufficient for determination of the taxonomic position of cyanobacteria by means of mathematical statistics. On the base of the linear discriminant analysis obtained spectroscopic data for 21 cyanobacterial strains from *CALU* collection were analyzed. It was shown that the presented technique allows an accurate differentiation of cyanobacteria up to the species/

Since the early 1950s, three different methods are commonly used to characterize phytoplankton and cyanobacterial samples taxonomically: high performance liquid chromatography (HPLC) [13–15, 23, 93]; flow cytometry [10, 11, 94, 95]; and optical microscopy. Various methods have been developed with the aims of increasing accuracy and yielding qualitative information. However, all of them have different limitations. Till now, the best taxonomic differentiation is still obtained using classical inverted microscopy. Unfortunately, this method is time-consuming, human-based and requires appropriate technical skills, and this eliminates the possibility of its application for continuous on-line monitoring. Nearly single-cell flow cytometric analysis is based on light scattering by the cells and fluorescence of the chlorophylls and the phycobilins. It can be easily automated, but it is appropriate only

whole cyanobacterial cultures.

group of unicellular cyanobacteria.

**spectra**

28 Cyanobacteria

Analysis of the *in-vivo* absorption and fluorescence spectra is an alternative way of obtaining qualitative information about the phytoplankton abundance and composition, which is continuously demonstrated by various publications [21, 23–27, 36, 97–100]. The relative phytoplankton abundance can be calculated once initial assumptions about the phytoplankton classes present and their pigment compositions have been made [10, 24, 25, 36, 100]. However, the correct classification of cyanobacterial species on the base of their fluorescence signature was hampered by alterations in pigment composition within one strain, which depends on the environmental conditions [93]. On the other hand, several researchers show that the nutrient and light limitations do not significantly change the initial fluorescence spectra and cannot impede the species discrimination [98, 101].

May be the first attempt to use phycoerythrins as chemotaxonomic markers was done by Glazer et al. [96] for red algae in 1982, but until now fluorescence spectra of phycobilins do not appear to be useful at familial, ordinal and class levels in taxonomic studies. Although the investigation in [96] concerns only purified high molecular weight phycoerythrin from red algae this work clearly demonstrates the possibility of the correct taxonomic analysis on the base of phycobiliproteins structural differences, which can serve as intrinsical fingerprints for taxons and genera in phytoplankton diversity. Later the correlation between the distribution of the biliproteins and the genera of *Cryptophyceae* was discussed in [102]. In 1985, Yentsch and Phinney [19] proposed an ataxonomic technique that utilized the spectral fluorescence signatures of major ocean phytoplankton. Seppälä and Olli [97] used spectral fluorescence signals to detect changes in the phytoplankton community. In 2002, Beutler et al. reported a reduced model of the fluorescence from the cyanobacterial photosynthetic apparatus designed for the in-situ detection of cyanobacteria and presented a commercially available diveable instrument for on-line monitoring of phytoplankton structure [21].

We elaborate a strict procedure for recording and processing single-cell fluorescence emission spectra, which eliminates the most of mentioned above difficulties and has a quite high classification accuracy. As well as according to our technique the fluorescence spectroscopic information is obtained via CLSM, the initial data has less variations and can be accurately sorted. Any objectionable and unpredictable impact can be eliminated at the first step of obtaining fluorescence spectra. Since noninvasive and nondestructive method is used the information about vital cell operation (e.g., light harvesting) can be additionally taken into account. All this allows one to obtain the desirable result directly following the procedure.

The classification procedure consists of three steps: (1) obtaining single-cell fluorescence spectra and creation of reference database; (2) data processing and extraction of classification parameters; (3) statistical analysis and evaluation of classification procedures.

wavelengths, that excite mostly one pigment (514, 543, and 633 nm), give less information than other laser lines and can cause more damage due to over-excitation. So, such excitation wavelengths should be used at the end of the record. To create a reference database of fluorescence spectra only the cells in normal physiological state should be used (if the study of the

Fluorescence Microscopic Spectroscopy for Investigation and Monitoring of Biological Diversity…

http://dx.doi.org/10.5772/intechopen.78044

31

If the database is formed from the cultured species, it is desirable to obtain reference spectra several times at different days and for various developmental stages of the culture to exclude any discrepancy and to take in account all possible variations in spectrum shape. The experimental sampling for each strain should include the sets of fluorescence spectra for more than 30–50 cells, to evaluate the statistical analysis. For the cultured species or for the strains from culture collection specified nutrient, temperature and light conditions should be applied, identical for all samples involved in classification. This is strongly required to exclude any adaptation effects. If the investigation is conducted over natural samples (for environmental monitoring), the reference database should be recorded for each tested reservoir because the difference in nutrient and light conditions could change the initial fluorescence spectra considerably. Moreover, this database should be extended by new experimental data constantly. While the reference

The whole procedure of obtaining intrinsic single-cell fluorescence spectra used in this study was designed to minimize preparatory manipulation, so as to conduct a noninvasive investigation of small amounts of experimental material and to prevent any damage of living cells. **Step 2.** While as the initial database is completed, the extraction and selection of classification parameters is carried out. To extract from the initial single-call fluorescence spectra, a set of classification parameters recently a computer program in the MATLAB software application [103] was elaborated. By means of this program, interpolation and smoothing of the raw spectra were carried out to eliminate the random noise and metering fluctuations. All spectra were reduced to the same scale and size of data array by the procedure of normalization and extrapolation. The first derivative was taken over smoothed and interpolated spectra and the fast Fourier transform was performed to exclude random noise, owing to the low intensity of the exciting and emitting energy. Else, some specific values characterizing the shape of the curves and the spectral composition of their derivatives were calculated, such as asymmetry and excess.

In **Figure 8**, several plots are given to illustrate the process of extracting parameters. At the left panel, normalized characteristic *in-vivo* single-cell fluorescence emission spectra at excitation wavelength 488 nm for two cyanobacterial species *Leptolyngbia CALU 1713* (blue lines) and *Nostoc CALU 1817* (red lines) are presented. Shaded regions mark the ranges of averaging for corresponding fluorescence intensities (575–586; 654–658; 679–685; and 714–723 nm). While getting the mean values at the corresponding area, one can calculate the impact of each region into the whole fluorescence intensity in percents. This will be the first set of classification parameters named fluorescence emission percentage contribution for individual pigments. For eight laser lines and for four zones as a result, we have a set of 32 parameters. The usual parameters for characterizing the shape of the curves asymmetry and excess (AE)—give another 16 parameters. Finally, the last set—the FFT-contributions in three specified regions of wavevector domain—is formed analyzing corresponding Fourier-transforms for each first derivation

depressed physiological state is not a case of current investigation).

spectra are available, the routine environmental monitoring is acceptable.

**Step 1.** To illustrate the usual form of spectra in Ref. database, in **Figure 7** several characteristic sets of single-cell fluorescence spectra are presented. Here four sets related to four cyanobacterial strains are shown: *Microcystis CALU 398*, *Merismopedia CALU 666*, *Leptolyngbya CALU 1715* and *Phormidium CALU 624*. Cyanobacterial strains are labeled according to *CALU* collection of the Core Facility Center for Culture Collection of Microorganisms of Saint-Petersburg State University. Each spectrum in the set was obtained by means of CLSM Leica TCS-SP5, using corresponding laser-line for excitation (405, 458, 476, 488, 496, 514, 543, and 633 nm). Corresponding excitation wavelengths are given over each spectrum. All spectra are normalized to the maximum intensity and shifted along x-axis for the clarity of observation. Four characteristic wavelengths, corresponding to the fluorescence maximum of different pigments can be easily distinguished at each spectrum: (1) peak near 580 nm corresponds to the fluorescence of phycoerythrin (is absent for *Microcystis* and *Leptolyngbya*), (2) peak near 656–560 nm corresponds to the fluorescence of phycocyanin and allophycocyanin in common (they are undistinguishable at room temperature), (3) peak near 682 nm corresponds to the fluorescence of chlorophyll *a* of PSII, and (4) peak or shoulder near 715 nm represents the fluorescence from PSI.

To obtain the representative fluorescent signature for given cyanobacterial strain by means of CLSM several points should be kept in mind. Each set of fluorescent spectra for single cell includes 4–8 spectra, obtained at different excitation wavelengths. One or two spectra in series is not enough for further differentiation. Thus, the low power settings should be used at all laser lines in order to eliminate cell damage during the record. Moreover, the excitation

**Figure 7.** Four characteristic sets of single-cell fluorescence spectra, corresponding to unicellular and filamentous cyanobacterial strains. The excitation wavelengths (405, 458, 476, 488, 496, 514, 543, and 633 nm) are given over the curves. All spectra are normalized to the maximum intensity and shifted along x-axis for convenience of observation.

wavelengths, that excite mostly one pigment (514, 543, and 633 nm), give less information than other laser lines and can cause more damage due to over-excitation. So, such excitation wavelengths should be used at the end of the record. To create a reference database of fluorescence spectra only the cells in normal physiological state should be used (if the study of the depressed physiological state is not a case of current investigation).

If the database is formed from the cultured species, it is desirable to obtain reference spectra several times at different days and for various developmental stages of the culture to exclude any discrepancy and to take in account all possible variations in spectrum shape. The experimental sampling for each strain should include the sets of fluorescence spectra for more than 30–50 cells, to evaluate the statistical analysis. For the cultured species or for the strains from culture collection specified nutrient, temperature and light conditions should be applied, identical for all samples involved in classification. This is strongly required to exclude any adaptation effects.

If the investigation is conducted over natural samples (for environmental monitoring), the reference database should be recorded for each tested reservoir because the difference in nutrient and light conditions could change the initial fluorescence spectra considerably. Moreover, this database should be extended by new experimental data constantly. While the reference spectra are available, the routine environmental monitoring is acceptable.

The whole procedure of obtaining intrinsic single-cell fluorescence spectra used in this study was designed to minimize preparatory manipulation, so as to conduct a noninvasive investigation of small amounts of experimental material and to prevent any damage of living cells.

**Step 2.** While as the initial database is completed, the extraction and selection of classification parameters is carried out. To extract from the initial single-call fluorescence spectra, a set of classification parameters recently a computer program in the MATLAB software application [103] was elaborated. By means of this program, interpolation and smoothing of the raw spectra were carried out to eliminate the random noise and metering fluctuations. All spectra were reduced to the same scale and size of data array by the procedure of normalization and extrapolation. The first derivative was taken over smoothed and interpolated spectra and the fast Fourier transform was performed to exclude random noise, owing to the low intensity of the exciting and emitting energy. Else, some specific values characterizing the shape of the curves and the spectral composition of their derivatives were calculated, such as asymmetry and excess.

In **Figure 8**, several plots are given to illustrate the process of extracting parameters. At the left panel, normalized characteristic *in-vivo* single-cell fluorescence emission spectra at excitation wavelength 488 nm for two cyanobacterial species *Leptolyngbia CALU 1713* (blue lines) and *Nostoc CALU 1817* (red lines) are presented. Shaded regions mark the ranges of averaging for corresponding fluorescence intensities (575–586; 654–658; 679–685; and 714–723 nm). While getting the mean values at the corresponding area, one can calculate the impact of each region into the whole fluorescence intensity in percents. This will be the first set of classification parameters named fluorescence emission percentage contribution for individual pigments. For eight laser lines and for four zones as a result, we have a set of 32 parameters. The usual parameters for characterizing the shape of the curves asymmetry and excess (AE)—give another 16 parameters. Finally, the last set—the FFT-contributions in three specified regions of wavevector domain—is formed analyzing corresponding Fourier-transforms for each first derivation

**Figure 7.** Four characteristic sets of single-cell fluorescence spectra, corresponding to unicellular and filamentous cyanobacterial strains. The excitation wavelengths (405, 458, 476, 488, 496, 514, 543, and 633 nm) are given over the curves. All spectra are normalized to the maximum intensity and shifted along x-axis for convenience of observation.

The classification procedure consists of three steps: (1) obtaining single-cell fluorescence spectra and creation of reference database; (2) data processing and extraction of classification

**Step 1.** To illustrate the usual form of spectra in Ref. database, in **Figure 7** several characteristic sets of single-cell fluorescence spectra are presented. Here four sets related to four cyanobacterial strains are shown: *Microcystis CALU 398*, *Merismopedia CALU 666*, *Leptolyngbya CALU 1715* and *Phormidium CALU 624*. Cyanobacterial strains are labeled according to *CALU* collection of the Core Facility Center for Culture Collection of Microorganisms of Saint-Petersburg State University. Each spectrum in the set was obtained by means of CLSM Leica TCS-SP5, using corresponding laser-line for excitation (405, 458, 476, 488, 496, 514, 543, and 633 nm). Corresponding excitation wavelengths are given over each spectrum. All spectra are normalized to the maximum intensity and shifted along x-axis for the clarity of observation. Four characteristic wavelengths, corresponding to the fluorescence maximum of different pigments can be easily distinguished at each spectrum: (1) peak near 580 nm corresponds to the fluorescence of phycoerythrin (is absent for *Microcystis* and *Leptolyngbya*), (2) peak near 656–560 nm corresponds to the fluorescence of phycocyanin and allophycocyanin in common (they are undistinguishable at room temperature), (3) peak near 682 nm corresponds to the fluorescence of chlorophyll *a* of PSII, and (4) peak or shoulder near 715 nm represents the

To obtain the representative fluorescent signature for given cyanobacterial strain by means of CLSM several points should be kept in mind. Each set of fluorescent spectra for single cell includes 4–8 spectra, obtained at different excitation wavelengths. One or two spectra in series is not enough for further differentiation. Thus, the low power settings should be used at all laser lines in order to eliminate cell damage during the record. Moreover, the excitation

parameters; (3) statistical analysis and evaluation of classification procedures.

fluorescence from PSI.

30 Cyanobacteria

**Figure 8.** The illustration for the classification parameter's calculation. Left panel: normalized in-vivo single-cell fluorescence emission spectra of two representative cyanobacterial species: blue lines—*Leptolyngbia CALU 1713*, red lines—*Nostoc CALU 1817*. Right panel: Fourier transformants normalized at maximum for the corresponding first derivations of the initial fluorescence spectra. In the inset the first derivation curves are plotted. Excitation wavelength 488 nm. Shaded regions show the bands of averaging.

plot (see **Figure 8**, left panel). The inset in **Figure 8** shows the first derivation curves for corresponding spectra. Three regions (43–58 μm−1; 95–110 μm−1; and 123–135 μm−1) were chosen and the mean value inside each was calculated. Thus, the last set of the parameters includes 24 values for each observation. Finally, for each spectra set, we extract 72 parameters, which are quite enough for evaluating classification by means of the linear discriminant analysis.

It should be noted that this procedure of the parameter's extraction varies according to the obtained data. The set of extracted parameters presented here is only an example of successful solution of formulated classification problem. Moreover, after statistical data analysis by means of hierarchical cluster analysis and stepwise linear discriminant analysis, the extracted set was reduced to 57 items to exclude a strict correlations between the parameters.

panel, the corresponding scaled regions are presented. In the legend, all used cyanobacterial strains are named and enumerated according to *CALU* collection. Solid curves bounded the regions, occupied by specified species: G—*Geitlerinema*, L—*Leptolyngbya*, O—*Oscillatoria*, M—*Microcystis*, Me—*Merismopedia*, P—*Pleurocapsa*, Pl—*Plectonema*, N—*Nostoc*, S—*Spirulina*,

**Figure 9.** The results of linear discriminant analysis. Solid curves bounded the regions, occupied by specified species: G—*Geitlerinema*, L—*Leptolyngbya*, O—*Oscillatoria*, M—*Microcystis*, Me—*Merismopedia*, P—*Pleurocapsa*, Pl—*Plectonema*, N—*Nostoc*, S—*Spirulina*, C—*Chlorogloea*, Sy—*Synechococcus*. The lower panel shows the scaled view of the corresponding

Fluorescence Microscopic Spectroscopy for Investigation and Monitoring of Biological Diversity…

http://dx.doi.org/10.5772/intechopen.78044

33

The classification accuracy in the presented example is near 98.3%. However, for the limited set of laser lines (4 instead of 7) and classification parameters (24 instead of 57), the classification accuracy does not reduce considerably—93.7%. The high classification accuracy is due to the fact that LDA works with distribution functions for classification parameters and their

It should be noted here that the limited set of laser lines and classification parameters was considered subjecting to further possible application of the presented technique in the microelectronic device for on-line data collection in environmental monitoring. During this investigation, the possibility of the implementation of data collection and data processing in one software-hardware application device on the base of on-a-chip technology was examined. Obviously, for the production of such device, it is quite desirable to have less radiation emis-

statistical characteristics, which allows to build a good classification model.

sion sources and limited number of detection spectral ranges.

C—*Chlorogloea*, Sy—*Synechococcus*.

regions.

**Step 3.** The last step includes the evaluation of classification procedure. Here we apply linear discriminant analysis (LDA), which is well known, and often applied to various biological objects. The procedure involved creating linear combinations of parameters with normal errors that best discriminate between site groups of cyanobacteria defined *a priori*. LDA was performed with the computer program designed in the MATLAB software [103], in which combinations of initial parameters were selected to maximize the ratio of group means discriminant scores to within-group variance [99, 104].

For this investigation, 314 sets of 8 spectra corresponding to 21 strains and 15 genera of cyanobacteria from the *CALU* collection were used. The results of Fisher discriminant analysis evaluated over 57 parameters is presented in **Figure 9**. The upper panel show 3D-plots in the space of discriminating functions. It is clear that the discrimination between species is sufficiently good. Moreover, the closely related species (e.g., *Spirulina* and *Oscillatoria*, *Synechococcus* and *Chlorogloea*, *Microcystis* and *Myxosarcina*) appear close to each other. Such species as *Leptolyngbia*, *Geitleninema* and *Oscillatoria*, which includes several strains, form a big groups. However, inside these groups, single strains also can be discriminated. In the lower Fluorescence Microscopic Spectroscopy for Investigation and Monitoring of Biological Diversity… http://dx.doi.org/10.5772/intechopen.78044 33

**Figure 8.** The illustration for the classification parameter's calculation. Left panel: normalized in-vivo single-cell fluorescence emission spectra of two representative cyanobacterial species: blue lines—*Leptolyngbia CALU 1713*, red lines—*Nostoc CALU 1817*. Right panel: Fourier transformants normalized at maximum for the corresponding first derivations of the initial fluorescence spectra. In the inset the first derivation curves are plotted. Excitation wavelength

plot (see **Figure 8**, left panel). The inset in **Figure 8** shows the first derivation curves for corresponding spectra. Three regions (43–58 μm−1; 95–110 μm−1; and 123–135 μm−1) were chosen and the mean value inside each was calculated. Thus, the last set of the parameters includes 24 values for each observation. Finally, for each spectra set, we extract 72 parameters, which are

It should be noted that this procedure of the parameter's extraction varies according to the obtained data. The set of extracted parameters presented here is only an example of successful solution of formulated classification problem. Moreover, after statistical data analysis by means of hierarchical cluster analysis and stepwise linear discriminant analysis, the extracted

**Step 3.** The last step includes the evaluation of classification procedure. Here we apply linear discriminant analysis (LDA), which is well known, and often applied to various biological objects. The procedure involved creating linear combinations of parameters with normal errors that best discriminate between site groups of cyanobacteria defined *a priori*. LDA was performed with the computer program designed in the MATLAB software [103], in which combinations of initial parameters were selected to maximize the ratio of group means dis-

For this investigation, 314 sets of 8 spectra corresponding to 21 strains and 15 genera of cyanobacteria from the *CALU* collection were used. The results of Fisher discriminant analysis evaluated over 57 parameters is presented in **Figure 9**. The upper panel show 3D-plots in the space of discriminating functions. It is clear that the discrimination between species is sufficiently good. Moreover, the closely related species (e.g., *Spirulina* and *Oscillatoria*, *Synechococcus* and *Chlorogloea*, *Microcystis* and *Myxosarcina*) appear close to each other. Such species as *Leptolyngbia*, *Geitleninema* and *Oscillatoria*, which includes several strains, form a big groups. However, inside these groups, single strains also can be discriminated. In the lower

quite enough for evaluating classification by means of the linear discriminant analysis.

set was reduced to 57 items to exclude a strict correlations between the parameters.

488 nm. Shaded regions show the bands of averaging.

32 Cyanobacteria

criminant scores to within-group variance [99, 104].

**Figure 9.** The results of linear discriminant analysis. Solid curves bounded the regions, occupied by specified species: G—*Geitlerinema*, L—*Leptolyngbya*, O—*Oscillatoria*, M—*Microcystis*, Me—*Merismopedia*, P—*Pleurocapsa*, Pl—*Plectonema*, N—*Nostoc*, S—*Spirulina*, C—*Chlorogloea*, Sy—*Synechococcus*. The lower panel shows the scaled view of the corresponding regions.

panel, the corresponding scaled regions are presented. In the legend, all used cyanobacterial strains are named and enumerated according to *CALU* collection. Solid curves bounded the regions, occupied by specified species: G—*Geitlerinema*, L—*Leptolyngbya*, O—*Oscillatoria*, M—*Microcystis*, Me—*Merismopedia*, P—*Pleurocapsa*, Pl—*Plectonema*, N—*Nostoc*, S—*Spirulina*, C—*Chlorogloea*, Sy—*Synechococcus*.

The classification accuracy in the presented example is near 98.3%. However, for the limited set of laser lines (4 instead of 7) and classification parameters (24 instead of 57), the classification accuracy does not reduce considerably—93.7%. The high classification accuracy is due to the fact that LDA works with distribution functions for classification parameters and their statistical characteristics, which allows to build a good classification model.

It should be noted here that the limited set of laser lines and classification parameters was considered subjecting to further possible application of the presented technique in the microelectronic device for on-line data collection in environmental monitoring. During this investigation, the possibility of the implementation of data collection and data processing in one software-hardware application device on the base of on-a-chip technology was examined. Obviously, for the production of such device, it is quite desirable to have less radiation emission sources and limited number of detection spectral ranges.

The universality of the considered technique makes it possible to use it for investigation of any phytoplankton species irrespective of their habitat or cultivation. Utilizing data from several fluorescence spectra, instead of one, results in more fingerprint information which leads to the taxonomic differentiation on a finer scale. Classification procedure, presented here, was carried out by means of statistical analysis on the base of mathematical characteristics of intrinsic fluorescence spectra of living single cells; therefore, it is free from usual subjectivity, which can occur while using methods of direct optical microscopy. Moreover, formalization of data processing gives a wide opportunity for automating of the classification procedure of cyanobacterial strains in field samples, while on-line monitoring of water bodies is conducted.

cope with the environmentally caused variations of the cyanobacterial fluorescence spectra. Moreover, an additional fluorescence information on the physiological state of cyanobacterial cultures provides a new information for predictive modeling and aquatic management,

Fluorescence Microscopic Spectroscopy for Investigation and Monitoring of Biological Diversity…

http://dx.doi.org/10.5772/intechopen.78044

35

The formalization of the genera identification and cultural physiological state analysis give an opportunity to develop a compact on-a-chip nanoelectronic device for preliminary on-line investigation of the field samples in situ and in vivo and for controlling of the laboratory

Obviously, the proposed methods require further development, including evaluation of more species representing more phytoplankton classes, and including non-taxonomic features, such as photoadaptation. Although the quantitative measurements were not performed in this study, they could be possible while all stages will be standardized. However, this work already demonstrates a high potential of fluorescence microscopic spectroscopy. We suggest that CLSM methods have potential application for several of the approaches noted earlier and also other studies regarding photosynthetic apparatus of cyanobacteria. We hope that the unique cell-biology of cyanobacteria will encourage further investigations because of their

[1] Abed RM, Dobretsov S, Sudesh K. Applications of cyanobacteria in biotechnology. Journal of Applied Microbiology. 2009;**106**(1):1-12. DOI: 10.1111/j.1365-2672.2008.03918.x

[2] Vijayakumar S, Menakha M. Pharmaceutical applications of cyanobacteria—A review.

[3] Singh S, Kate BN, Banerjee UC. Bioactive compounds from cyanobacteria and microalgae: An overview. Critical Reviews in Biotechnology. 2005;**25**(3):73-95. DOI: 10.1080/073

[4] Liu L. New Bioactive Secondary Metabolites from Cyanobacteria [Thesis]. University of

[5] Grewe CB, Pulz O. The biotechnology of cyanobacteria. In: Ecology of Cyanobacteria II. Netherlands: Springer; 2012. pp. 707-739. DOI: 10.1007/978-94-007-3855-3\_26

Journal of Acute Medicine. 2015;**5**(1):15-23. DOI: 10.1016/j.jacme.2015.02.004

alternatively to the delayed fluorescence described in [108].

growing importance in rural biotechnology and commercial production.

\* and Ludmila Chistyakova<sup>2</sup>

1 Saint-Petersburg Electrotechnical University, Saint-Petersburg, Russia

2 Saint-Petersburg State University, Saint-Petersburg, Russia

\*Address all correspondence to: renes3@mail.ru

cultures during industrial incubation.

**Author details**

Natalia Grigoryeva<sup>1</sup>

**References**

88550500248498

Helsinki: Helsinki; 2014

Undoubtedly, the data set should be expanded to include more species and phytoplankton classes/divisions, grown under different nutrient and light conditions. However, this work already demonstrates the potential of the discrimination of phytoplankton classes by fluorescence microscopic spectroscopy. This work lays the foundation for determining cyanobacterial abundance by direct fluorescence measurement of sea- and freshwater. Combining the knowledge of phytoplankton structure along with taxon-specific measurements of photosynthetic activity and biochemical cell composition, can lead to new models which increase the reliability of on-line monitoring.
