Aquatic Ecology 36: 165–178, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.

165

Vertical structure and photosynthetic activity of Lake Shira phytoplankton Nikolai A. Gaevsky1, Tatiana A. Zotina2 and Tamara B. Gorbaneva1 1 Krasnoyarsk 2 Institute

State University 660041 Svobodny av. 79, Krasnoyarsk, Russia (E-mail: [email protected]); of Biophysics of SB RAS, Akademgorodok, Krasnoyarsk, 660036, Russia

Accepted 10 August 2001

Key words: chlorophyll a, fluorescence, taxonomic structure, variable fluorescence

Abstract This study was conducted to analyse vertical dynamics of phytoplankton distribution in Shira Lake during the summer stratification regime. From late June to September phytoplankton in Shira Lake were stratified with the maximum in the lower part of the thermocline, at a depth of 8–12 m, with a chlorophyll concentration up to 23 µg and biomass up to 5 mg l−1 . Maxima of chlorophyll and biomass of cyanobacteria and green algae were in different layers. From June to September a major part of chlorophyll a was in green algae, while under ice – in cyanobacteria. The variable fluorescence proves high photosynthetic activity of algae in the depth assemblage. Epifluorescent analysis disclosed that additional light-harvesting pigments were better developed in cells from the depth maximum. The maximum of gross primary production calculated from fluorescence corresponded to the depth maximum of phytoplankton. Primary production over a season was 2.7 gO2 m−2 . Formation mechanisms of the depth maximum of phytoplankton are discussed in this paper.

Introduction Shira Lake is a brackish, thermally stratified lake with hydrogen sulphide in the hypolimnion. Lakes of this type are often characterised by a stratified distribution of phytoplankton within thermally stable layers (Reynolds, 1992). The lake remains among little investigated ecosystems. The purpose of our work is to describe seasonal dynamics of phytoplankton and their major taxonomic groups and to study photosynthesis and photosynthetic activity of microalgae. The data will be used to discuss possible mechanisms responsible for the temporal and spatial structure of phytoplankton in Shira Lake and to verify a mathematical model of the ecosystem. This work is a step towards the main aim of our study: to reveal interrelations between vertical structure, biomass, composition of phytoplankton, and input of nutrients to Shira Lake. Taxonomic and functional characteristics of phytoplankton were studied from 1998 to 2000 by fluorescence methods based on works by Kiefer (1973),

Yentsch & Phinney (1985), Bates (1985), Samuelsson & Öquist (1977), Kuladaivelu & Daniell (1980), Vincent (1981), Falkowski & Kiefer (1985). Chlorophyll fluorescence is widely used to study spatial and temporal distribution of phytoplankton (Yentsch & Phinney, 1985; Wilhelm et al., 1987). Yentsch & Yentsch (1979) and Yentsch & Phinney (1985) applied the fluorescence method to determine the taxonomic structure of a microalgal community.

Methods Sampling Water samples were taken in the central part of the lake with a 1-l Ruthner sampler or a 4-l Molchanov sampler. After sampling, water samples were protected from light, kept in a cold place and processed in two hours.

166 Temperature measurement Temperature was measured with either a mercury thermometer mounted in the inner volume of the water sampler or a submersible sensor (thermal resistor built into an electron circuit with a digital indicator). In both cases the measurement accuracy was ± 0.2 ◦ C. Light attenuation In situ attenuation of solar radiation (W m−2 ) was measured with a portable underwater irradiance meter, fabricated at Krasnoyarsk State University. The light sensor included a silicon diode and a glass filter with passband 540 ± 30 nm. It was intercalibrated with a standard meteorological Yanishevsky pyronometer (Russia). The sensitivity of the submerged sensor made it possible to evaluate optical characteristics of the water column to the depth of 12 – 13 m. The Secchi depth transparency was measured with a 30-cm white disk.

Figure 1. Transmission (T) spectra of light filters. Fluorescence excitation: blue-green light filter No. 22 (1); blue light filter No.15 and blue-green light filter No. 22 (2); yellow-green light filter No. 9 and blue-green light filter No. 9 (3); orange light filter No. 11 and blue-green light filter No. 22 (4). Fluorescence emission: red light filter No. 18 and purple light filter No. 8 (5).

Fluorescence Fluorescence was measured with a single-beam single-cell fluorometer devised and fabricated at Krasnoyarsk State University. The basic specifications of the instrument were as follows: the cell volume was 27 cm3 ; the magnetic mixer operated at 100 rpm. The source of unmodulated exciting light was a halogen incandescent lamp (100 W). The exciting light was separated into spectral ranges by changing the combinations of light filters. The characteristics of the spectral domains were as follows: A was a broad optical band 400–620 nm with six intensity steps (from 24 to 474 µmol photons m−2 s−1 ); B was a narrow band with the maximum at 410 nm; C was a narrow band with the maximum at 510 nm and D was a narrow band with the maximum at 540 nm. Fluorescence emission was recorded through a dark red filter and a purple light filter (> 680 nm). Spectral characteristics of excitation domains and domains of fluorescence recording are shown in Figure 1. After a cell was filled with a subsample of phytoplankton, the fluorescence signals were recorded in the following sequence: 1 – steady-state levels (Fss ) with broad-band light excitation (see (i) A of previous paragraph) from the lower intensity towards the higher; 2 – maximum levels (Fm ) after addition of 10 µM DCMU [3-(3,4-dichlorophenyl)-1,1dimethylurea] in the same order; 3 – maximum levels (Fm ) excited by blue (410 nm), blue-green (510 nm) and green (540 nm)

light. Background fluorescence in water samples filtered under vacuum through a 0.75–0.85µ-pore-sized membrane and standard fluorescence emission (Fst ) of the deep red light glass filter were measured after excitation by the same light filters as in i.3. For each case net fluorescence (Fn ) was calculated as follows: 1 , (1) Fn = (Fph − Fb ) Fst where Fph is the fluorescence of a sample, Fb is the background fluorescence, Fst is the fluorescence emission of the deep red light glass filter. The net signals of cyanobacteria (Cyan) and green (Chlor) algae were determined by the difference in the net fluorescence of the sample after excitation by blue-green (F 510) and green (F 540) lights. Their evaluation was based on the solution of a system of linear equations derived from the condition that the net fluorescence signal recorded in phytoplankton Fn at a prescribed wavelength was the sum of net signals of cyanobacteria (Cyan) and green algae (Chlor). F (510)n = F (510)Cyan + F (510)Chlor, (2) F (540)n = F (540)Cyan + F (540)Chlor. Given the ratio of fluorescence signals F (510)Cyan/ F (540)Cyan is kCyan , and F (510)Chlor/F (540)Chlor is kChlor , then the system of linear equations can be solved with respect to F (510)Cyan and F (510)Chlor as follows:

167 F (510)Cyan =

kCyan [F (510)n − kChlorF (540)n ] , kCyan − kChlor

F (510)Chlor = F (510)n − F (510)Cyan.

(3)

The previous study of individual green algae and cyanobacteria cells by epifluorescent analysis showed that the yielded average kCyan = 0.5 and kChlor = 2.0. The calculated values of net fluorescence signals F (510)Cyan and F (510)Chlor and the chlorophyll a concentration values measured spectrophotometrically in the same samples CChl(a) (after ethanol extraction) were used as a basis for the equation: CChl(a) = K1 F (510)Cyan + k2 F (510)Chlor.

(4)

Values of coefficients k1 and k2 were found by MANOVA: k1 = 29.6 ± 5.9 (p-level 0.00001) and k2 = 20.1 ± 3.1 (p-level 0.00001). Epifluorescence (λmax ∼ 685 nm) in individual algal cells was recorded with a LUMAM-E microscope (LOMO, Russia). Epifluorescence was excited with a 70 W halogen lamp. Spectral sections of excitation were distinguished with a red separating plate in the microscope and one of three combinations of ‘colour’ light filters, whose characteristics are given above (Figure 1). To produce statistically significant results the fluorescence parameters were measured for 10–20 cells of each species for each sample. Potential photochemical activity of photosystem 2 (PS2) was evaluated by the variable fluorescence induced by addition of DCMU, an electron transport inhibitor, into the suspension of algae (Öquist et al., 1982; Keller, 1987). Variable fluorescence was calculated by the formula: Fm − Fss . (5) Fvar = Fm Gross primary production (GPP) (gO2 m−3 h−1 ) was calculated by the empirical equation (Gaevsky et al., 2000): GPP = b(I )Fvar Cchl I,

(6)

where b(I) is an experimentally evaluated function of the light intensity, I, given by: b(I ) = 6.277 × 10−3 I −0.5 for I > 4 W m−2 and b(I ) = 3 × 10−3 for I < 4 W m−2 , Fvar is the variable fluorescence (see Equation (5)), Cchl is the chlorophyll a concentration (mg m−3 ), and I is the light intensity (W m−2 ) in the layer where production was evaluated.

Phytoplankton counting For microalgal counting, subsamples were concentrated on Vladipor (Russia) membranes (pore size 0.85–0.95 µm) under low vacuum and preserved with Kuzmin’s reagent (10 g KI, 50 ml H2 O, 5 g I, 5 ml chromic acid (1%), 10 ml acetic acid, 80 ml formalin (40%) (Kuzmin, 1975). Microalgae were identified and counted at ×400 or ×1000 magnification with phase contrast under a Karl Zeiss microscope in a Fusch–Rosenthal chamber (V = 3.2 mm3 ). The size of algal cells was measured and wet biomass of algae was calculated according to the volumes of similar geometric shapes. Chlorophyll extraction For chlorophyll extraction water samples were concentrated under vacuum on Vladipor (Russia) 0.95– 1.05 µm-pore-sized membranes covered with a thin layer (1–2 mm) of powdered MgCO3 . The volume of a water subsample passed through one filter was 100–200 ml. Before extraction, filters were dried at room temperature. Chlorophyll was extracted in 90% ethanol according to Nusch (1980). Layers of MgCO3 containing chlorophyll were removed from filters, placed in ethanol and heated at 78 ◦ C for not more than 1 min and then extracted for 24 hours at room temperature. After that MgCO3 was removed by centrifuging. Samples were acidified with 2 M HCl (Nusch, 1980). Extinction of extracts was measured in a UVIKON 943 double beam UV/VIS spectrophotometer (Kontron, France).

Results Site description During the summer period of June–September the lake is thermally stratified. The epilimnion is the layer 6 m from the surface (temperature 19–22 ◦ C). The thermocline is at a depth of 6–8 m where the temperature jump is up to 12 ◦ C. From the depth of 12–13 m to the bottom the temperature changes from 2.5 to 1.2 ◦ C. During the stratification regime the oxygen maximum is at a depth of 6–8 m; the oxygen and hydrogen sulphide zones in the lake intersect at a depth of 11 to 12 m (Kalacheva et al., 2002).

168 Light attenuation and euphotic zone boundaries During the period of summer-autumn stratification (July–September), the vertical extinction coefficient for 540 nm (ε540 m−1 ) was relatively stable in the layer from 0 to 6.5 m (0.122 ± 0.001 m−1 ) and nonmonotonically increased in the 7–12 m layer (0.289 ± 0.008 m−1 ) (Figure 2). The boundary of the euphotic zone corresponding to 1% of the surface light intensity is at a depth of 11 meters. At low temperatures in deep layers of Shira Lake the euphotic zone boundary can decrease to 0.1%, at a depth of 14–14.5 m. Secchi disk transparency in the central part of the lake during the summer stratification period was 3.75 to 4.0 m. The average extinction coefficient of green light was 0.46 to 0.49 of Secchi disk transparency. Separation of fluorescence signal Earlier works (Zotina et al., 1999) demonstrated that during the summer period Shira Lake had a stratified distribution of the total phytoplankton biomass and the biomass of all major dominants (except diatoms) with a maximum in the methalimnion at a depth between 8 and 15 m. The dominant species were Lyngbya contorta Lemm., Microcystis sp. Kutz., and Dictyospaerium tetrachotomum Printz. In summer, diatoms developed in the epilimnion, with Cyclotella tuberculata Makar et Log dominating. In March, Nitzschia palea W.Sm. and Navicula lanceolata (Ag.) Kutz. prevailed in the lower edge of ice. In this case, i.e. when plankton are represented by green algae and cyanobacteria, a taxonomic index to separate the chlorophyll of cyanobacteria from that of green algae is the ratio of net fluorescence signals F510 /F540 , excited by blue-green (510 nm) and green (540 nm) light. As quantitative equivalents of chlorophyll a we use net fluorescence signals of green algae and cyanobacteria excited by blue-green (510 nm) light (Equation 3). The coefficient of correlation between chlorophyll a concentrations calculated from fluorescence and those found by extraction was r = 0.67, and the coefficient of correlation between the total chlorophyll a concentration calculated by fluorescence and the total cell biomass estimated by cell count was r = 0.84 (n = 32). The chlorophyll a concentration of green algae (Cchlor = k2 F (510)chlor) and cyanobacteria (Ccyan = k1 F (510)cyan) calculated from chlorophyll fluorescence correlated with the respective biomass values (rchlor = 0.72, n =

32; rcyan = 0.63, n = 22). Hence it follows that the fluorescent method makes it possible to produce an adequate pattern of spatial distribution of the two major taxa of microalgae. Spatial and temporal distribution of phytoplankton Fluorescence The first observations were made on 13 March 1999, when the lake surface was covered with ice (1.2 to 1.3 m thick). Profiles of chlorophyll a of green algae and cyanobacteria are shown in Figure 3. The chlorophyll a concentration in the layer from 0 to 13 m was 1.8 to 2.2 µg l−1 and decreased in the near-bottom layer. Up to 80% of chlorophyll a was accounted for by cyanobacteria. L. contorta was the dominant phytoplankton species under ice. In June (19 June 1999) vertical distribution of algae became inhomogeneous (Figure 3). The green algae had a maximum chlorophyll concentration in the layer from 0 to 8 m. In the lower strata the community was mixed. In 1999 and 2000 the vertical profiles of chlorophyll were similar, however, in 1999 the malfunctioning fluorometer yielded inhomogeneities in the vertical profiles that were not confirmed by the microscopically defined phytoplankton distribution. Therefore, we discuss the results obtained in July–September 2000. Observations made from early July to mid-September showed a vertical distribution with low chlorophyll a concentrations in the layer from 0 to 6 m and a maximum at 8 to 13 m (Figure 3). The main peak of green algae was above the main peak of cyanobacteria. Absolute chlorophyll a concentrations for green algae were higher than for cyanobacteria. During summer the main peak of chlorophyll concentration gradually moved deeper and became narrower. In November, immediately before freeze-up, vertical distribution of pigment in the 0 to 11 m layer became homogeneous (1.5–2.0 µg l−1 in green algae and 1 µg l−1 in cyanobacteria). In the near-bottom layer, between 14 and 15 m, the chlorophyll a concentration of green algae dropped almost to zero, while that of cyanobacteria increased to 3 µg l−1 (Figure 3). Differences in the depth of water in the pelagic zone did not affect the shape of the chlorophyll a distribution from the surface to the bottom (Figure 4). Distribution of algal biomass Vertical distribution of phytoplankton in June–August 1999 in the pelagic zone of the lake is shown in Fig-

169

Figure 2. Vertical profile of light transmission (open squares) and extinction coefficient (open circles) for 540 nm radiation (ε540 m−1 ) 12 September 2000. SD – Secchi depth.

ure 5. In late June and in July the maximum of total algal biomass was at a depth of 8 m, in August it was at a depth between 10 and 12 m. The biomass of algae at the depth maximum increased during the season from 3 to 5 mg l−1 . By the end of June the vertical distribution of green algae was uniform; with time, the biomass formed its peak in July at a depth of 6 to 8 m and at 10 m in August (Figure 5). In June–August Botryococcus braunii Kutz. (Chlorophyta) generally has several outbreaks in the surface layer. On 19 June the biomass of this species at the surface was 28 mg l−1 (Figure 5). The biomass of Botryococcus was transported over the lake surface by wind currents and finally piled up near the shore. Therefore, this species was neglected when we investigated the formation of the depth maximum of phytoplankton in the pelagic zone of the lake. Unlike the vertical distribution of green algae, which is largely formed by one dominant, D. tetrachotomum, the profile of cyanobacteria is formed by two species, L. contorta and Microcystis sp. The maximum of biomass of L. contorta was at a lower depth than the maximum of Microcystis sp. and D. tetrachotomum (Figure 5). During the season the maximum of L. contorta was observed at a depth between 6 and 10 m. Microcystis spp. had its biomass peak at a depth between 8 and 12 m. In July L. contorta formed an additional biomass peak at a depth of 13 m, with biomass 0.6 mg l−1 . At a depth of 16 m there was a peak of Synechocystis sp., with biomass 0.5 mg l−1 (Figure 5). Throughout the season the maxima of the dominants remained in their specific layers. Even when the bio-

mass assemblages descended, the maximum of every species retained its position (Figure 5). In June–August 1999 the total biomass under a square meter of the water column, from the surface to the bottom, was 27 to 48 g m−2 (Table 1). In the upper 12-m layer the biomass of phytoplankton comprised more than 70% of the biomass in the entire water column (Table 1). According to our calculations cyanobacteria dominated in the total biomass, amounting to 34–64%, green algae constituted 36– 65%, and the portion of diatoms in some periods reached 22%. L. contorta and D. tetrachotomum made approximately equal contributions to the total biomass: 25–44% and 18–32%, respectively. In the upper 12-m layer the ratios of algal taxa were close to those in the entire water column (Table 1). Pigment modification To investigate the pigment composition of the phytoplankton dominant species, the cells of D. tetrachotomum, L. contorta and Microcystis sp. were collected from different depths and analysed by the epifluorescent method. This method made it possible to characterise the pigments of algae by their spectrum of fluorescence excited by blue (410 nm), bluegreen (510 nm), and green (540 nm) spectral bands. To characterise the fluorescence action spectrum we used not the absolute signals, but ratios of signals: F510 /F410 , F540 /F410 , F510 /F540 (Figure 6). The two species of cyanobacteria had much larger F510 /F410 and F540 /F410 ratios in the 12–16 m layer than in the

170

Figure 3. Vertical profiles of chlorophyll a measured by fluorescence.

171

Figure 4. Vertical profiles of total chlorophyll a calculated from chlorophyll fluorescence in three pelagic stations of different depths on 25 June 1999. Table 1. Biomass and proportional composition of phytoplankton per one square meter during summer stratification Date

18 June 1999 16 July 1999 21 July 1999 6 Aug. 1999

Total biomass, (g m−2 )

Cyclotella

Part of the total biomass, % Lyngbya Microcystis Dictyosphaerium

1

2

1

2

1

2

1

2

1

2

48.1 29.2 26.9 27.9

43.4 25.8 22.4 20.7

1.8 2.6 3.1 19.3

1.8 2.5 3.4 21.8

24.6 37.5 43.8 27.3

21.6 35.9 41.9 23.9

9.0 22.5 16.0 14.9

7.7 22.1 15.0 13.2

18.1 29.9 32.1 22.3

17.8 31.2 34.5 21.2

1 – In the whole water column (0–20 or 22 m). 2 – In the 0–12 m layer.

layers above and below. The F510 /F540 ratio was relatively stable throughout the water column. Conversely, the peak of the F510 /F540 ratio of D. tetrachotomum was at a depth of 12 m. Photosynthetic activity of algae – variable fluorescence As an index of photosynthetic activity of algae we have chosen variable fluorescence (Fvar ), which indicates real photosynthetic efficiency (according to the standard nomenclature of fluorescence characteristics and quenching coefficients) (van Kooten & Snel, 1990). Its values are closest to the maximum efficiency of excitation energy capture at a low intensity of the exciting light (Krause & Laasch, 1987). The minimal intensity of the exciting light of the broad spectral band was 24 µmol photons m−2 s−1 , or 6 W m−2 . Vertical profiles of Fvar recorded under these conditions around noon showed variations in this parameter depending on depth (Figure 7). The index was

markedly lower in the surface layer and had two peaks at depths of 2–4 and 10 m. In the near-bottom layers Fvar dropped almost to zero. In August 1998, we investigated diurnal variations of Fvar and found that the parameter slightly decreased by the end of the day in samples taken from the top layers. The maximum of variable fluorescence in the vertical profile was in the layer of 8 to 12 m, Fvar = 0.4 to 0.5. At the same time, variable fluorescence of algae lifted from a certain layer decreased with an increase in the exciting light intensity. The relation of the Fvar value to depth (H ) and light intensity (I ) was approximated by equation: Fvar = 0.207 − 0.661H + 0.151I + 0.728H 0.97 − 0.159I 0.992. Gross primary production of phytoplankton was determined at the central station in clear windless weather (12.09.00). Primary production was calculated by empirical equation (6).

172

Figure 5. Vertical profiles of algae dominating in phytoplankton.

173

Figure 6. Parameters of epifluorescence excitation spectra in phytoplankton cells collected on 25 July 2001. Fluorescence signal ratio – F 510/F410 (squares), F 540/F410 (triangles), F 510/F540 (diamonds).

174

Figure 7. Vertical profile of variable fluorescence in August 2000; smoothing curve derived by distance weighted least squared ± eps.

Light intensity (I ) in the layer of GPP determination was calculated using surface light intensity (I0 ) and light transmittance of the water column (Table 2). Light intensity in the PAR range (I0 ) was recorded at 0.5- to 1-h intervals between 8 and 14 h (half a daytime). Chlorophyll concentration was measured by the fluorescent method described above (Table 2). Thus, in the calculation of the gross primary production for different times of the day in every study layer the only constants were chlorophyll concentration and light transmittance. Then the GPP values collected between 8 and 14 h were integrated and multiplied by 2, resulting in the GPP for the daytime. The vertical profile of gross primary production is shown in Figure 8.

Table 2. Data for calculation of gross primary production (12 September 2000): light transmittance (ε540 ) and chlorophyll concentration Depth (m)

ε540

Chl a, (µg l−1

1 2 4 6 8 9 10 11 12 13 14

1 0.536 0.322 0.187 0.076 0.037 0.019 0.008 0.003 0.002 0.001

1.1 1.0 1.1 1.0 4.0 9.4 20.2 20.8 23.5 3.9 5.1

Discussion From Figure 3 it is apparent that the depth maximum of the chlorophyll a concentration above 10 µg l−1 forms by the beginning of July. During summer the main peak of chlorophyll concentration gradually moves deeper and becomes narrower. In July– September, when green algae (Figure 3) are not recorded deeper than 15 m, the chlorophyll a concentration reaches 19 µg l−1 . The chlorophyll of cyanobacteria can be found as deep as 20 m, and its maximum does not exceed 10 µg l−1 . The shapes of profiles retrieved by fluorescence are in good agreement with the vertical profiles of total biomass. However, the employment of the fluorescent method alone results in the loss of information on species specificity

of chlorophyll and the spatial structure of the dominant species. There are several well-known reasons for the formation of the phytoplanktom depth maximum. Sedimentation of algae is among them (Lindholm, 1992; Reynolds, 1992). During sedimentation cells either die or adapt to changing light conditions. The species of Lyngbya are capable of forming deep phytoplankton maxima (Reynolds et al., 1983; Trifonova, 1990). This species does not form gas vacuoles. L. contorta, a dominant species in Shira Lake, is regarded as a low light adapted species (Havens et al., 1998). A greater amplitude of the Microcystis sp. biomass peak

175

Figure 8. Gross primary production profile (GPP), derived by Equation (6), experiment of 12 September 2000.

in the deep maximum of this species suggests that the peak has been formed due to biomass increment. This somewhat contradicts the general knowledge about the ecology of the genus Microcystis and calls for further research. Algae and cyanobacteria manifest their adaptation to altered light conditions in the deep layers by the development of additional pigments. Potential existence of several spectral forms sometimes interferes with the analysis of pigment variation in the light-harvesting complex of PS2 using spectrophotometry (fluorometry) of pigments in vivo. Nevertheless, the F510 /F410 and F540 /F410 ratios of cyanobacteria can correlate with the ratio between the phycobilin complex and chlorophyll a in PS2, and F510 /F540 – with the ratio between the short-wave phycoerythrin and the phycocyanine of longer wavelength (Bertrand & Schoefs, 1997). In green algae the F510 /F410 ratio can correspond to the ratio of chlorophyll a to chlorophyll b and F540 /F410 and F510 /F540 can represent the correlation between yellow pigments and chlorophyll a. Increasing ratios of F510 /F410 and F540 /F410 , with little change in F510 /F540 , in L. contorta and Microcystis sp. found in deep layers (Figure 6) indicate the development of phycobilin light-harvesting complex without apparent domination of either phytoerythrin or phycocyanin (Figure 6). The increase in the D. tetrachotomum F510 /F540 ratio with a peak at a depth of 12 m also indicates development of the lightharvesting complex equally by chlorophyll a and b (as the F510 /F410 ratio is stable). Development of the PS2 light-harvesting complex in cyanobacteria and green algae at the interface of photic and aphotic zones must

be due to the replacement of nutrient limitation in the photic zone by light limitation in the aphotic zone. The existence of such an interaction between phosphorus and light in the regulation of phytoplankton photosynthesis was described by Millard et al. (1996). It is well known that the phycobilin complex can be rapidly degraded under nutrient deficiency. Under conditions of macronutrient limitation, many cyanobacteria degrade their phycobilisomes in a rapid and orderly fashion (Grossman et al., 1993). A possible reduction of the phycobilin complex in layers above 12 m (Figure 6) should manifest itself in a considerably lower chlorophyll/biomass ratio in cyanobacteria in the upper layers. This can be among the reasons for the lack of strict conformity of the vertical chlorophyll profiles (Figure 3) defined by fluorescence to the algal biomass evaluated by direct microscopy (Figure 5). The real efficiency of photosynthesis (Figure 7) was found to decrease with an increase in the exciting light intensity. Photosynthesis of phytoplankton in the upper layer of Shira Lake may be inhibited by light. The highest values of Fvar were observed in the depth maximum (Figure 7). A decrease in variable fluorescence in the layer below 15 m can be due to several reasons. First is sedimentation of ‘old’, photosynthetically inactive, cells. The second, and, as we think, more important, reason is increasing concentration of dissolved hydrogen sulphide. The addition of 0.075 mg l−1 H2 S had no marked effect on the kinetics of Scenedesmus variable fluorescence (Brack & Frank, 1998), but at a concentration of 2.6 mg l−1 of H2 S it was almost completely inhibited. Hence, H2 S should

176 have an inhibiting effect on green algae at a depth of 14 m. Vertical distribution of GPP in Shira Lake (Figure 8) demonstrates not only the depth maximum of phytoplankton biomass but also the depth maximum of primary production. These data agree with the data based on photoassimilation of 14 C (Kopylov et al., 2002). Daily production can be estimated as 3.1 gO2 m−2 for specific conditions of 12.09.00 and as 4.85 gO2 m−2 for modelled conditions of midJuly. From June to September vertical distribution of light and chlorophyll a is fairly stable. The number of clear days is almost similar to that of overcast days when intensity drops by an order of magnitude. Therefore, average gross daily primary production over a season can be evaluated as (4.85 ± 0.485)/2 = 2.7 gO2 m−2 , with a active vegetation period of 120 days GPP=320 gO2 m−2 , or 4545 J m−2 a year. Regarding production characteristics Shira Lake can be compared to Chedenyarvi Lake in Karelia (4400 J m−2 yr−1 ), Krasnoye Lake on the Karel neck (3700 J m−2 yr−1 ) and classified as a mesotrophic water body (S < 5 m) (Buillion, 1983). Anoxygenic photosynthesis might be among the reasons why the cyanobacteria form a depth maximum at the boundary of the hydrogen sulphide zone. The measurements of light attenuation showed that the photic zone intersects the hydrogen sulphide zone. In the layer about 13 m, purple bacteria have been observed to develop in summer. There are grounds to assume that the cyanobacteria forming assemblages at the boundary of the photic zone in the presence of H2 S (L. contorta; Microcystis spp., Synechocystis sp.) survive owing to their capacity not only to retain viability in the presence of hydrogen sulphide, but also to use it as an electron donor in anoxygenic photosynthesis. Certain cyanobacteria are known to change from oxygenic photosynthesis to anoxygenic and back (Fenchel & Riedl, 1970), e.g., Oscillatoria limnetica, phylogenetically related to the genus Lyngbya (Giovannoni et al., 1988; Cohen et al., 1975). As photosystem 2 is not needed to oxidise hydrogen sulphide in anaerobic conditions (Fry et al., 1984; Shahak et al., 1987), transition of algae to anoxygenic photosynthesis should be accompanied by the disappearance of DCMU- stimulated variable fluorescence, which has been observed in Shira Lake (Figure 7). The phycobilin light-harvesting antenna related to PS2 might be reduced under these conditions (Campbell et al., 1998); this is in good agreement with a decrease in F510 /F410 and F540 /F410 ratios of spectral bands

of fluorescence excitation in the near-bottom stratum (Figure 6). Certain cyanobacteria are known to be capable of mixotrophic and heterotrophic growth. Karnaukhov (1988) showed that colonial and filamentous cyanobacteria, Microcystis pulverea and Phormidium uncynatum, can, under certain conditions, contain both chlorophyll-rich cells and cells that have partly or completely lost chlorophyll. The author came to a conclusion that the former actively photosynthesise, releasing oxygen, while the nutrition of the latter is heterotrophic and they consume oxygen. This property allows the algae to survive in the absence of oxygen where other algal species die. Our assumption that cyanobacteria dominate in Shira Lake because of anoxygenic photosynthesis calls for experimental validation by isotope methods. The ability of heterotrophic and mixotrophic growth invites further investigation of changes in the cell microstructure (Martinez et al., 1991; Orus & Martinez, 1991), of changes in the pigment ratio (Karnaukhov, 1988) or of some other parameters. It has been found that vertical distribution of phytoplankton is in antiphase with that of zooplankton (Zotina et al., 1999). In contrast with phytoplankton, the maximum of zooplankton is in the epilimnion. This means that algae are most intensively consumed in the epilimnion. Besides, it is common knowledge that the algae in the well-heated epilimnion are consumed by zooplankton at maximum rates, as the filtration rate of most zooplankters reaches its maximum at a temperature of 15–20 ◦ C (Gutelmacher, 1986). Hence, grazing can prevent development of algae in the epilimnion and indirectly facilitate development of the depth maximum of cyanobacteria and phototrophic sulphur bacteria. Higher transparency of the top layers allows penetration of more light into the deeper layers. In addition, turbulent diffusion delivers nutrients from the hypolimnion. Still, it is not quite clear how the green algae form their maximum in the metalimnion. Apparently, peculiar vertical distribution of nutrients in Shira Lake makes development of algae in the lower part of the metalimnion possible. Kalacheva et al. (2002) showed that in summer the maximum phosphorus and ammonium concentrations are in the hypolimnion. At the same time, in the photic zone concentrations of these elements are low and are often below the detection level. The observations may suggest that at the beginning of summer a major part of phytoplankton in Shira Lake occurs in the epiliminion, at the upper boundary of

177 the photic layer, but at the end of summer – at the lower boundary of the photic layer and hypolimnion. At the beginning of summer the biomass in the epilimnion is, probably, formed by the growth of green algae and at the end of summer by the growth of cyanobacteria. It can be assumed that the metabolic pathway that determines ecosystem functioning also changes from algal-based to bacteria-based. The cause of the ‘switch’ can be the onset of nutrient limitation of algae in the upper, well-illuminated, layer. Outbreaks of Botryococcus and Cyclotella also facilitate the onset of nutrient limitation in the epilimnion. So, as a result of stratification the epilimnion community of phytolankton is only indirectly related with the metalimnion.

Acknowledgements Authors are grateful to Dr Andrei G. Degermendzhy for his support of this study, to Elena S. Kravchuk for assistance in carrying out measurements and to anonymous reviewers for thoughtful and useful comments. The work has been done with support of grants: INTAS No 97-0519, RFBR No 99-05-64333?, Federal Special-Purpose Program ‘Integratziya’ No 73.

References Bates SS (1985) Sample preconditioning for measurement of fluorescence induction of chlorophyll a in marine phytoplankton. J Plankt Res 7: 703–714 Bertrand M and Schoefs B (1997) Working with photosynthetic pigments: problems and precautions. In: Pessarakli M (ed.), Handbook of Photosynthesis. Marcel Dekker, New York, pp. 151–172 Brack W and Frank H (1998) Chlorophyll a fluorescence: a tool for the investigation of toxic effects in the photosynthetic apparatus. Ecotoxicol Environ Safety 40: 34–41 Buillion VV (1983) Primary Production of Plankton in Inland Water Bodies. Nauka, Leningrad Campbell DC, Hurry V, Clarke AK, Gustafson P and Öquist G (1998) Chlorophyll fluorescence analysis of cyanobacterial photosynthesis and acclimation. Microbiol. Molec Biol Rev 63: 667–683 Cohen Y, Jorgensen BB, Padan E and Shilo M (1975) Sulphidedependent anoxigenic photosynthesis in the cyanobacterium Oscillatoria limnetica. Nature 257: 489–492 Falkowski P and Kiefer DA (1985) Chlorophyll a fluorescence in phytoplankton: relationship to photosynthesis and biomass. J Plankt Res 7: 715–731 Fenchel TM and Riedl RJ (1970) The sulfide system: a new biotic community underneath the oxidized layer of marine sand bottoms. Marine Biol 7: 225–268 Fry I, Robinson AE, Spath S and Packer L (1984) The role of Na2 S in anoxygenic photosynthesis and H2 production in

the cyanobacterium Nostoc muscorum. Biochem Biophys Res Commun123: 1138–1143 Gaevsky NA, Kolmakov VI, Popelnitsky VA, Gold VM and Dubovskaya OP (2000) Light effect factor in fluorescent evaluation of photosynthesis intensity in plankton microalgae. Fiziologiya Rastenii 47: 930–935 Giovannoni SJ, Turner S, Olsen GJ, Barns S, Lane DJ and Pace NR (1988) Evolutionary relationships among cyanobacteria and green chloroplasts. J Bacteriol 170: 3584–3592 Grossman AR, Schaefer MR, Chiang GG and Collier JL (1993) The phycobilisome, a light-harvesting complex responsive to environmental conditions. Microbiol Rev 57: 725–749 Gutelmacher BL (1986) Metabolism of Plankton as a Single Whole: Trophometabolic Interactions of Zoo- and Phytoplankton. Nauka, Leningrad Havens KE, Phlips EJ, Cicha MF and Li B-L (1998) Light availability as a possible regulator of cyanobacteria species compositionin a shallow subtropical lake. Freshw Biol 39: 547–556 Kalacheva GS, Gubanov VG, Gribovskaya IV, Gladchenko IA, Zinenko GK and Savitsky SV (2002) Chemical analysis of Lake Shira water (1997–2000). Aquat Ecol 36: 123–141 Karnaukhov VN (1988) Spectral Analysis of Cells in Ecology and Environment Protection. Department for scientific and technical information at the Scientific Center for Biological Studies of the USSR Academy of Sciences, Pushchino Keller A (1987) Mesocosm studies of DCMU-enhanced fluorescence as measure of phytoplankton photosynthesis. Mar Biol 96: 107–114 Kiefer DA (1973) Fluorescence properties of natural phytoplankton population. Mar Biol 22: 263–269 Kopylov AI, Kosolapov DB, Degermendzhy NN, Zotina TA and Romanenko AV (2002) Phytoplankton, bacterial production and protozoan bacterivory in stratified, brackish-water Lake Shira (Khakasia, Siberia). Aquat Ecol 36: 205–218 Krause GH and Laasch H (1987) Energy-dependent chlorophyll fluorescence quenching in chloroplasts correlated with quantum yield of photosynthesis. Z Naturforsch 42: 581–584 Kuladaivelu G and Daniell H (1980) Dichlorophenyl dimethylurea (DCMU) induced increases in chlorophyll a fluorescence intensity – an index of photosynthetic oxygen evolution in leaves, chloroplasts and algae. Physiol Plant 48: 385–388 Kuzmin GV (1975) Species composition and abundance of phytoplankton. In: Mordukhai-Boltovsky FD (ed.), Methods for Investigation of Inland Aquatic Biogeocenoses. Nauka, Moscow, pp. 73–87 Lindholm T (1992) Ecological role of depth maxima of phytoplankton. Arch Hydrobiol Beih 35: 33–45. Martinez F, Ascaso C and Orus MI (1991) Morphometric and stereologic analysis of Chlorella vulgaris under heterotrophic growth conditions. Ann Bot 67: 239–245 Millard ES, Myles DD, Johannsson OE and Ralph KM (1996) Seasonal phosphorus deficiency of lake Ontario phytoplankton in two index stations: light versus phosphorus limitation of growth. Can J Fish Aquat Sci 53: 1112–1124 Nusch EA (1980) Comparison of different methods for chlorophyll and phaeopigment determination. Arch Hydrobiol Beih 14: 14– 36 Öquist G, Hagstrom A, Alm P, Samuelson G and Richardson K (1982) Fluorescence as an alternative method for estimating primary production. Mar Biol 68: 71–75 Orus MI and F Martinez (1991) Chlorophyll a/b ratio and thylacoid stacking modification in Chlorella vulgaris UAM 101. Biochem Physiol Pflanzen 187: 197–202

178 Reynolds CS (1992) Dynamic, selection and composition of phytoplankton in relation to vertical structure in lakes. Arch Hydrobiol Beih 35: 13–31 Reynolds CS, Tundisi JG and K Hino (1983) Observations on a metalimnetic Lyngbya population in a stable stratified tropical lake (Lagoa Carioca, Eastern Brasil). Arch Hydrobiol 97 (1): 7–17 Samuelsson G and Öquist G (1977) A method for studying photosynthetic capacity of unicellular algae based on in vivo chlorophyll fluorescence. Physiol Plant 40: 315–319 Shahak Y, Arieli B, Binder B and Padan E (1987) Sulfide-dependent photosynthetic electron flow coupled to proton translocation in thylakoids of the cyanobacterium Oscillatoria limnetica. Arch Biochem Biophys 259: 605–615 Trifonova IS (1990) Ecology and Succession of Lake Phytoplankton. Nauka, Leningrad van Kooten O and Snel JFH (1990) The use of chlorophyll fluorescence nomenclature in plant stress physiology. Photosynth Res 25: 147–150

Vincent WF (1981) Photosynthetic capacity measured by DCMUinduced chlorophyll fluorescence in an oligotrophic lake. Freshw Biol 11: 61–78 Wilhelm C, Kramer P and I Wiedemann (1987) Die Lichtsammelkomplexe der verschiedenen Algenstamme. Biologie in Unserer Zeit 17(5): 138–143 Yentsch CS and Phinney DA (1985) Spectral fluorescence: a taxanomic tool for studying the structure of phytoplankton populations. J Plankton Res 7: 617–632 Yentsch CS and Yentsch CM (1979) Fluorescence spectral signatures: the characterization of phytoplankton populations by the use of excitation and emission spectra. J Mar Res 37: 471–483 Zotina TA, Tolomeyev AP and Degermendzhy NN (1999) Lake Shira, a Siberian salt lake: ecosystem structure and function. 1 Major physico-chemical and biological features. Int J Salt Lake Res 8: 211–232