Phytoplankton species succession in a shallow Mediterranean lake (L. Kastoria, Greece): steady-state dominance of Limnothrix redekei, Microcystis aeruginosa and Cylindrospermopsis raciborskii
Maria Moustaka–Gouni1*, Elisabeth Vardaka1# Eleni Tryfon1&
1 Department of Botany, School of Biology, Aristotle University of Thessaloniki, GR-541 24 Thessaloniki, Greece
# Current address: Department of Fisheries and Aquaculture Technology, Alexander Technological Educational Institute of Thessaloniki, Campus of Nea Moudania, P.O. Box 157, GR-632 00 Nea Moudania, Greece
Current address: Administration of Environmental Planning, Hellenic Ministry for the Environment, Physical Planning and Public Works, GR-112 51 Athens, Greece
*Corresponding author ()
Key words: phytoplankton succession, functional groups, cyanobacteria steady-states, polymictic Mediterranean lake
This paper has not been submitted elsewhere in identical or similar form, nor will it be during the first three months after its submission to Hydrobiologia
Abstract
The phytoplankton species composition and seasonal succession were examined in Lake Kastoria during the period November 1998 – October 1999. A total of 67 species and 19 functional groups were identified. Only four out of the 67 species, all Cyanobacteria, were dominant (Limnothrix redekei, Microcystis aeruginosa, Cylindrospermopsis raciborskii and Aphanizomenon gracile). Diatoms were rare, not only in terms of species number, but also in terms of biomass (contributing <5% to the total phytoplankton biomass) in relation to the rather low silicon concentrations throughout the year. The functional groups S1, SN, M and H1 were found dominant in the lake. The species A. gracile (functional group H1) behaved like the species Cylindrispermopsis raciborskii (functional group SN) which is tolerant to mixing and poor light conditions. The phytoplankton seasonal succession showed similar patterns in all six sampling stations, both at the surface and the bottom water layer, with minor differences during Microcystis aeruginosa dominance. Two steady-state phases were identified within a year lasting for 4 months under relatively stable physical conditions. In these steady-states, the Limnothrix redekei persistent dominance under low light availability and low inorganic nitrogen has been explained by its specific ability such as buoyancy regulation to exploit resources in the water column. Moreover, high population densities over the winter and before the development of daphnids may contribute to the steady-state dominance of Limnothrix. Different niches separated vertically in the water column is one of the explanations for the Limnothrix-Microcystis steady-state when a replacement between the two species was observed in different water layers and areas of the lake. Long lasting steady-states of Cyanobacteria observed in Lake Kastoria and in other Mediterranean and tropical freshwaters may indicate influence of warm climate properties on phytoplankton dynamics.
Introduction
Phytoplankton dynamics has been the subject of many studies in freshwaters but only in a few cases it is examined in relation to equilibrium/non-equilibrium theories (e.g. Salmaso, 2003). For identification of equilibrium states in phytoplankton seasonal succession, Sommer et al. (1993) set three criteria: (i) a maximum of 3 species of algae contribute more than 80% of total biomass, (ii) their dominance persists for more than 1-2 weeks and (iii) during that period the total biomass does not increase significantly. Recent studies dealing with “equilibrial” species and assemblages clarify our understanding of steady-states in phytoplankton succession in a wide spectrum of freshwaters, mostly from mid-latitudes (Naselli-Flores et al., 2003). Different types of phytoplankton steady-states have been explained as the result not necessarily of competition but due to several other processes (e.g. grazing, species specific abilities; Albay & Akcaalan, 2003; Rojo & Alvarez-Cobelas, 2003).
Steady-state phases are rarely attained in phytoplankton succession. However they have been observed more regularly in shallow hypertrophic lakes where Cyanobacteria are primarily the protagonists (Mischke & Nixdorf, 2003; Nixdorf et al., 2003). A Limnothrix redekei steady-state assemblage has been reported so far only in one case (Rojo & Alvarez Cobelas, 2003). L. redekei, a typical phytoplankter in turbid mixed layers (Reynolds et al., 2002) of lakes and lowland rivers of central and northern Europe (Meffert, 1989), is not common in southern Europe (Gkelis et al., 2005). In contrast, Cylindrospermopsis raciborskii is a low-latitude species known for its invasive behavior in mid-latitudes (Padisák, 1997). C. raciborskii, Aphanizomenon gracile and Microcystis aeruginosa have been found dominant in summer steady-state assemblages of hypertrophic shallow wetlands in southern Europe (Stoyneva, 2003).
In this work, we examine the seasonal succession of phytoplankton during an annual study in Lake Kastoria, a highly eutrophic shallow lake. The phytoplankton species are classified according to functional groups proposed by Reynolds et al. (2002) and Padisák et al. (2003). When examining succession, we try to identify steady-state phases and to understand the environmental factors that promote, maintain and disturb the dominant species in these phases. This investigation on a relatively large shallow lake in the Mediterranean region will contribute to the limited knowledge of the compositional diversity of dominants in steady-state phases of shallow lakes, the frequency and longevity of the phases. Moreover, the study of Cyanobacteria steady-states may have application in the much needed measures to restore water quality in eutrophic lakes.
Study site
Lake Kastoria (Fig. 1) is situated at latitude 40ο30′ N, and longitude 21ο18′ E in Northern Greece. It covers 24 km2, has a maximum depth of 8 m, an average depth of 4 m and a water retention time greater than 2 years. The lake’s outflow discharge is man controlled through manipulation of the water level when it reaches its maximum. A substantial water inflow increase during January-March 1999 caused the lake to overflow and the local authorities discharged large amounts of water in the spring of that year in order to control the water level.
Human impact on the lake and its catchment was diverse. Hydraulic adjustments, fish stock management with the introduction of cyprinoids and macrophyte cutting are just some among them. Moreover, Kastoria is an urban lake that had been receiving sewage effluents for decades until 1995. Former studies of Lake Kastoria have been made that show among other things high concentrations of inorganic nitrogen and phosphorus (Moustaka-Gouni et al., 2006). Nevertheless, there are periods when inorganic nitrogen and phosphorus fall below the threshold values used to detect N and P limitation (Reynolds et al., 2002). Dissolved silicon never exceeded 15 μmol l-1 in the lake water (Table 1). The peak values of planktic Cyanobacteria biomass indicate a highly eutrophic system (Vardaka et al., 2000) that has a history of toxic cyanobacterial blooms (Lanaras et al., 1989; Cook et al., 2004). Possible effects of toxic cyanobacterial blooms on heterotrophic nanoplankton, both positive and negative, and a weak structure of microbial food web in the lake, have recently been reported (Moustaka-Gouni et al., 2006).
Methods
Sampling was carried out from November 1998 to October 1999 fortnightly during the warm period of the year and monthly during the cold period. Ice cover on the lake, thin in some periods, prohibited sampling in December 1998. Water samples were collected from six stations in the deeper area of the lake’s basin (S1, S2, S3, S4, S5, S6; Fig. 1). The samples were collected from the surface (0-1 m) and the bottom layer (one meter above sediment varying from 4 to 7 m at the maximum depth).
The methods used for in situ measurements, chemical analyses of nutrients and microscopical analysis of phytoplankton have been described by Moustaka-Gouni et al. (2006). Phytoplankton counts (cells, filaments, colonies) were performed using the inverted microscope method. To convert colony counts of Microcystis aeruginosa to cell numbers, the average number of cells of 30 colonies was determined using the equation of Reynolds & Jaworski (1978). Cell and filament volumes were estimated from appropriate geometric formulae after measuring the dimensions of 30 cells/filaments.
The mixing zone (zmix) was identified using temperature profiles and the euphotic zone (zeu) calculated as 2.0 times the Secchi depth. An index of light availability (LI) was calculated according to Makulla & Sommer (1993):
LI = 2(SD/ zmix) x (D/24)
where LI is the light index, SD is Secchi depth (m), zmix is mixing depth and D is daylength (h).
Phytoplankton functional groups were established according to Reynolds et al. (2002) and Padisák et al. (2003). Species were considered dominant if they contributed more than 10% to the total phytoplankton biomass in each individual sampling date. Steady-state phases (SSI, SSII) were identified when (i) 1, 2 or 3 phytoplankton species contributed more than 80% to total biomass, (ii) their existence or co-existence occurred for more than 3 weeks and (iii) during that period the species composition of the community was almost unchanged and the total phytoplankton biomass differed non significantly (ANOVA, p>0.05) between sampling dates or less than 20% from that of the previous sampling date value.
A one way ANOVA test was used to compare the means of phytoplankton biomass among sampling dates. A Pearson correlation analysis was used to determine relationships between biological variables. The relationship between the biomass of dominant species and temperature was analyzed using nonlinear regression analysis. Principal Component Analysis (PCA; Legendre & Legendre, 1998) was used to examine the relationship between physical and chemical properties of the water and the relative dominance of phytoplankton dominant species during the period of their coexistence. In PCA analysis and when necessary, log transformation of values were made to achieve normality.
Results
Phytoplankton species composition and biomass
A total of 67 phytoplankton species have been identified (Table 2) in the water samples examined throughout the year. Chlorophytes contributed the highest number of species (29) followed by Cyanobacteria (20), diatoms (5), dinophytes (4), cryptophytes (3), euglenophytes (2), xanthophytes (2), chrysophytes (1) and prymnesiophytes (1). The functional group J was the best represented in number of species followed by M and H1 (Table 2).
Phytoplankton biomass consisted mainly of Cyanobacteria (contributing 91.8% to the annual mean biomass) (Fig. 2). Abrupt seasonal variations in total phytoplankton biomass with almost constant contribution by different taxonomic groups have been observed throughout the year. Nevertheless, periods with rather constant species composition and biomass have also been recognized (phases SSI and SSII; Fig. 2).
Only 4 out of the 67 species were dominant throughout the year. These species, all Cyanobacteria, belong to different functional groups of phytoplankton: Limnothrix redekei to S1, Cylindrospermopsis raciborskii to SN, Microcystis aeruginosa to M and Aphanizomenon gracile to H1. The species Aphanizomenon issatschenkoi (H1), Ceratium hirundinella (LM), Cryptomonas sp. (Y), Peridinium sp. (LO), Monoraphidium griffithii (X1), Phacotus lenticularis (YPh) and Nitzschia acicularis (D) contributed <10% to the total phytoplankton biomass. The nanoflagellates Rhodomonas minuta (X2) and Chrysochromulina parva (X2), although abundant, contributed <5% to the total phytoplankton biomass.
The four dominant species made up 91% of mean total phytoplankton biomass influencing significantly the temporal distribution of the total phytoplankton biomass (r=0.989, p<0.001). Limnothrix redekei persisted throughout the year showing the highest biomass during the cold period of the year (Fig. 3A). Cylindrospermopsis raciborskii and Aphanizomenon gracile developed during the summer when the temperature exceeded 20 °C (Figs 3B, C). Microcystis aeruginosa population increased from late summer to autumn in the range of 17 - 27 °C (Fig. 3D).
PCA was used to examine how dominant species are grouped in relation to physical and chemical parameters during the summer-autumn period (Fig. 4) when all dominant species coexisted. The first two axes with the largest eigenvalues explain 37.7 and 29.5 % of the total variance in the data, respectively. Cylindrospermopsis highest dominance (>40%) is differentiated along axis I by lower nitrate and ammonia nitrogen concentrations, lower zmix:zeu and N:LI ratios. Along axis II, Cylindrospermopsis and Aphanizomenon dominance together with Limnothrix is differentiated by higher temperatures while Microcystis co-dominance with Limnothrix is differentiated by higher phosphate phosphorus concentrations, and higher zmix:zeu and N:LI ratios.
Phytoplankton seasonal succession
Similar patterns of phytoplankton succession have been observed in all six sampling stations both at the surface and the bottom layers (Fig. 5). However, the spatial similarity in phytoplankton dominance declined during August – October when Microcystis aeruginosa contributed >50% of the total phytoplankton biomass in the sampling stations S2 and S6. Two different types can be recognized in the course of phytoplankton succession: a) the mono-dominance of Limnothrix redekei and b) the co-dominance of i) L. redekei-Cylindrospermopsis raciborskii-Aphanizomenon gracile and ii) Limnothrix redekei-Microcystis aeruginosa-Cylindrospermopsis raciborskii.
The mono-dominance of Limnothrix was observed during both the high and low phytoplankton biomass periods (January – March and May – June, respectively). The first period is identified as a steady-state phase of phytoplankton (SSI; Fig. 5). This species contributing >80% to the total phytoplankton biomass persisted for 2 months, when composition and total biomass of phytoplankton did not change considerably (see Fig. 2). This steady-state phase was observed almost identically in different areas and layers of the lake (Fig. 5). The period of the steady-state phase is characterized by low water temperature (between 3.9 and 8.2 °C) and low light availability (LI ranged from 0.07 to 0.15) throughout the isothermally mixed column (Appendices 1, 2). Nitrate and ammonia nitrogen remained low (< 2.5 μmol l-1and <1.4 μmol l-1, respectively) and phosphate phosphorus ranged between 0.2 and 0.4 μmol l-1, (Table 1).
Over summer, three species of phytoplankton, Limnothrix redekei, Cylindrospermopsis raciborskii and Aphanizomenon gracile persisted for 2 months but, since total phytoplankton biomass changed considerably (see Fig. 2), no steady-state stage can be recognized. This period is characterized by intermittent thermal stratification (Appendix 2) and high light availability (up to 1.2; Appendix 1). Low N:P (min. 1.4) and N:LI (min. 8.1) resource ratios (Appendix 3) prevailed during the period of Cylindrospermopsis raciborskii and Aphanizomenon gracile dominance.
A second steady-state phase (SSII; Fig. 5) identified during the period August–October. The species Microcystis aeruginosa and Limnothrix redekei constituting 82% of phytoplankton biomass persisted for 3 weeks. For the next 3 weeks, Microcystis aeruginosa and Limnothrix redekei still co-dominated and with the minor contribution of Cylindrospermopsis raciborskii made up >80% of phytoplankton biomass. During the whole period (August – October), the species composition was almost unchanged and the total phytoplankton biomass did not change significantly (ANOVA, p>0.05). This phase was different in regard to the order of dominance (1st Microcystis, 2nd Limnothrix or replaced alternatively) in the different water layers and areas of the lake (Fig. 5). The period is characterized by low light availability (LI < 0.1) in warm mixed layers (Appendices 1, 2) and high concentrations of phosphate phosphorus (0. 6 to 1.0 μmol l-1), (Table 1).
Discussion
Phytoplankton flora of Lake Kastoria is poor in comparison to the rather rich flora of shallow eutrophic lakes of similar size in the area (Tryfon & Moustaka-Gouni, 1997; Temponeras et al., 2000). The low number of species may reflect the long lasting period of multiple human impacts on the lake (sewage effluents, hydrological adjustments, fish stock management) resulting in a strong environmental filter for phytoplankters. Loss of biodiversity and increase of algal blooms are the most evident negative ecological impacts of human activities on the microbial level in aquatic systems (Paerl et al., 2003). In addition, the low number of diatoms in the phytoplankton species list of this shallow lake, in relation with the rather low silicon concentrations throughout the year (Table 1), may indicate that past competition has at least partially shaped the phytoplankton species pool of the lake (see Sommer, 1990). Of the 33 functional groups described by Reynolds et al. (2002) and Padisák et al. (2003), 19 were represented in the lake’s phytoplankton.
Diatoms were rare, not only in terms of species number, but also in terms of biomass (max. 1.2 mg l-1, contributing <5% to the total phytoplankton biomass) (Fig. 2) even during the winter-spring season, the typical period of diatoms (Sommer et al., 1986). The low diatom biomass is related to low Si:P <62.0 and Si:N <25.0 resource ratios during winter (Appendix 4). Nitzschia acicularis dominated within the diatoms. However, the biomass of N. acicularis was only 1.4% of the total phytoplankton biomass, whereas that of Limnothrix redekei was 48 times higher. Both species grow well under low temperatures and low light availability in mixed layers. However, dominance of L. redekei has been found associated with anthropogenic eutrophic conditions in shallow lakes (Meffert, 1989), while Nitzschia acicularis perform better in not enriched and deeper mixed waters (Huszar et al., 2003).
Limnothrix redekei dominance exhibited very similar patterns in distinct sampling areas and water layers affecting considerably the patterns of phytoplankton succession. The similarity of succession patterns declined when Microcystis aeruginosa, contributing >50% to the total phytoplankton biomass, formed surface accumulations. Spatial differences (Fig. 5) may arise through the interaction of the dominants’ behavior (buoyancy regulation, auto-regulated surface accumulation) and the pelagic zone physical dynamics under medium speed local winds (physically induced accumulation).
Limnothrix redekei made up to 99% of the phytoplankton biomass in winter (Fig. 5) setting the diversity close to zero (Shannon index based on biomass Hb=0.2; Moustaka-Gouni, unpubl. data). An extremely low phytoplankton diversity (zero) has been reported, to the best of our knowledge, only in one case (Borics et al., 2000). The main factors that may have promoted and maintained the persistent steady-state of L. redekei, were relatively stable physical conditions (mixing at low temperatures) (Appendix 2) and rather constant a) low light conditions (Appendix 1) and low nutrient concentrations (Table 1), and b) low Si:P, Si:N resource ratios (Appendix 4) for its possible competitors. Under these conditions, the specific abilities of species, such as photoadaptation and buoyancy regulation in the water column (Reynolds et al., 2002), to effectively exploit resources, may have considerably contributed to the development of this steady-state. Enhanced cell gas-vacuolation in a large number of L. redekei trichomes (up to 50% cell volume; Gkelis et al., 2005), has been observed during the winter. Moreover, high population densities over the winter (up to 90x106 trichomes l-1), before the development of daphnids (Moustaka-Gouni et al., 2006), may be another key factor for establishing and maintaining overdominance of L. redekei (e.g. Nicklisch, 1999).