Journal of Plankton Research Vol.20 no.4 pp.691-708, 1998
The role of microcystins in heavy cyanobacterial bloom formation
Bojan Sedmak and Gorazd Kosi
National Institute of Biology, Department of Biology, Biotechnical Faculty,
University of Ljubljana, Vecnapot 111, 1000 Ljubljana, Slovenia
Abstract The presence of high microcystin concentrations in cyanobacterial blooms additionally
affects species diversity. Blooms with high toxin contents can reach higher cell densities, which is also
demonstrated by microcystin cell contents. In vitro experiments show that microcystins influence
phytoplankton proliferation. The action is strongly dependent on the phytoplankton species tested
and light conditions. We propose that the environmental impact of different microcystins depends on
their enzymatic inhibition activity and thus could not be measured merely on the basis of their toxicity
to vertebrate species. Their role in heavy cyanobacterial bloom and scum formation is discussed, as
well as their impact on the massive proliferation of other species following toxic cyanobacterial bloom
degradation.
Introduction
The majority of authors point out the significance of cyanobacteria as a health
hazard to humans and agricultural livestock (e.g. Codd, 1994; Falconer et ai,
1994). Nevertheless, we would like to emphasize the negative impact cyanobacterial blooms have on the aquatic environment. Besides the health hazard to
higher vertebrates, they have a great influence on general environmental health
and must be considered as a warning of environmental degradation. Eutrophication, together with blooms as the consequence, has a strong influence on all living
organisms in water bodies (e.g. Harper, 1992). There are a great number of
measurable physicochemical and environmental factors that are responsible for
phytoplankton distribution and species diversity, including bloom formation.
These factors can affect all the organisms in a water body or only specific, more
susceptible ones, having regard to the conditions, amounts or concentrations.
Variations in day length, insolation, temperature, wind mixing, rainfall, flushing
and nutrient loading rates affect all kinds of water bodies, but are more
pronounced in smaller ones. Physicochemical parameters vary more widely in
ponds and reservoirs than in deep natural lakes, and reflect more or less only
temporary conditions. As blooms are frequent in smaller water bodies in Slovenia
(Sedmak et al., 1994; Sedmak and Kosi, 1997), we focused our research mainly on
phytoplankton species abundance and composition as an indicator of environmental status. In our opinion, phytoplankton species are the best reflection of
long-lasting physicochemical and biological factors, which are themselves functions of the climatic and hydrological regimes, since the majority of freshwater
phytoplankton organisms incorporate resistant benthic resting stages into their
life history strategies. In this way, planktonic algae can compensate for noxious
exogenic short-term effects.
Despite the fact that cyanobacteria have lower maximal growth rates than the
majority of other phytoplankton species, in certain conditions they are able to
outstrip all other species, resulting in cyanobacterial blooms. Their chances of
© Oxford University Press
691
RSedmak and G-Kosi
proliferation are better in low-light conditions (e.g. Oscillatoria rubescens), as
demonstrated by Mur et al. (1978). These authors unsuccessfully tried to demonstrate that conditions of low phosphate concentration or other nutrients could be
the factor that allowed cyanobacterial dominance. Two main factors can generate low-light conditions: the depth of the water and its turbidity. The latter could
be caused by materials suspended in the water, such as mud, or, more commonly,
by phytoplankton. Light distribution in water is highly influenced by the mutual
shading of phytoplankton. In this way, the turbidity of the water increases and
favours those organisms adapted to low light intensities. Light limitation alone
does not elucidate the cyanobacterial bloom phenomenon. In a competition
experiment, Mur et al. (1978) presumed that there must be an ecological factor
which lowers the growth of algae more than the growth of cyanobacteria. We
would like to rephrase this idea as: there must be an ecological factor which
augments the growth of cyanobacteria more than the growth of algae.
It has long been known that the possibility of a cyanobacterial bloom being
toxic is >50% (Olson, 1964). In our investigations, >80% of Microcystis blooms
were toxic (Sedmak et al., 1994; Sedmak and Kosi, 1997). Additionally, a longterm study showed that each year Microcystis aeruginosa was non-toxic at the
beginning of the growing season, and developed high toxicity during the first
strong biomass increase (Benndorf and Henning, 1989). Although there are
several hypotheses elucidating the factors influencing cyanobacterial dominance
(e.g. Shapiro, 1990), none of them deals directly with the influence of microcystins
on bloom formation.
Cyanobacteria are most probably the first organisms in the Earth's history, with
fossil records dated 3.3-3.5 billion years ago, and therefore it is unlikely that
microcystins play the role of a defensive substance. We propose that microcystins
act as growth regulators, helping cyanobacteria to produce as many offspring as
possible, giving a better opportunity for successful adaptation. Our goal is to
emphasize the effects of microcystins on the phytoplankton community during
hepatotoxic blooms.
Method
Research sites
We analysed data from seven fish ponds and three reservoirs over a 3 year period
(Tables I and II). The locations were grouped according to outflow operativity
and hepatotoxin contents. Some data from Lake Bled (a natural lake) are also
discussed.
Field sampling
Qualitative samples were taken in late summer and autumn, during cyanobacterial blooms, with a 25 mesh net as a vertical profile and analysed for phytoplankton species community composition. The surface bloom was avoided in
order to achieve a representative sample of the community. Samples were
preserved in 5% formaldehyde. With the use of an inverted microscope, they
692
Microcystins and bloom formation
Table L Description of locations with relatively stable conditions during cyanobacterial blooms,
grouped according to cumulative hepatotoxin contents of the bloom
Locations without outflow
Sampling
Description
site
Year
Chi a
("gl-1)
Blooms with high microcystin contents (ZMC > 10 ug I"1)
BOC
Gravel pit
19%
67
HLP
Gravel pit/fish pond
1994
169
HGD
Fishpond
1994
249
SAV
Reservoir/fishpond
1994
91
Blooms with low microcystin contents (SMC < 10 ug I"1)
BOR
Gravel pit/
1994
31
fish pond
19%
11
LUT
Fishpond
1994
56
19%
40
PODC
Gravel pit/fish pond
1994
51
POD
Gravel pit/fish pond
19%
57
Phytoplankton
species no.a
Trophic category
(OECD values)
6(2)
6(2)
5(2)
8(5)
Hypertrophic
Hypertrophic
Hypertrophic
Hypertrophic
21 (16)
24 (19)
18(13)
20(15)
12(8)
19(14)
Hypertrophic
Hypertrophic
Hypertrophic
Hypertrophic
BOC, Boreci; BOR, Borovci: BUK, Lake Bukovnik; HGD, Hotinja Village (fish pond); HLP, Hotinja
Village (gravel pit); KOS, Koseze; LED, Ledava; LUT, Lutverci; POD, Podgrad; SAV, Savci.
'Figures in parentheses represent phytoplankton taxa other than cyanobacteria.
•"Because of low rainfall, the outflow was inoperative.
Intermediate value.
Table II. Description of locations with operational outflows resulting in unstable conditions during
cyanobacterial blooms with subsequent higher phytoplankton diversity
Locations with outflow
Sampling
Description
site
BUK
KOS
KOS
LED
SAV
Reservoir
Clay pit/fish pond
Clay pit/fish pond
Reservoir
Reservoir
Year
Chi a
(MgH)
Phytoplankton
species no.°
Trophic category
(OECD values)
1996
1994
11
49
20
31
30
15 (12)
18(16)
28(25)
24(22)
22 (17)
Eutrophic
Eutrophic
Eutrophic
Hypertrophic
Hypertrophic
19%
1995
1996
"Figures in parentheses represent phytoplankton taxa other than cyanobacteria.
were analysed for phytoplankton species composition and abundance. The
species were identified according to Komarek (1958,1991), Starmach (1966,1980)
and Hindak (1978).
Toxin analysis
Bloom samples were purified, concentrated and extracted according to the
method of Harada et al. (1988), as described elsewhere (Sedmak and Kosi,
1997). The toxic fractions were separated using HPLC. All equipment was
obtained from Waters, Millipore Division, and consisted of a Waters 600 Multisolvent Delivery System, a Waters 616 Pump, a Waters 996 Photodiode Array
Detector and a Waters Fraction Collector (Milford, MA) equipped with an
analytical Hibar Pre-Packed RT 125-4 (Merck) LiChrospher 100 RP-18 (5 um)
693
B^edmak and G.Kosi
column. The equipment was fed through an NEC Image 466 ES computer run
by a Millenium 2010 Chromatography Manager (Millipore). The amounts of
toxins were estimated by comparison of the peak area at 238 nm of the test
sample after separation with methanol:0.05 M phosphate buffer (58:42, pH 3.0)
with those of standard samples (MC-LR, ICN Biomedicals Inc.; MC-LR, MCRR and MC-YR, Calbiochem]. Microcystin isolation was performed under the
previously described conditions using a preparative Spherisorb S10 ODS2
column (Phase Separation Inc. USA) with a flow rate of 10 ml min"1.
Organisms and growth conditions
For in vitro experiments, non-axenic phytoplankton species isolated in our
laboratory were grown in continuous culture. Twelve hour illumination was
provided by Osram L 18 W/72 Biolux white fluorescent lamps together with
Sylvania Gro-Lux F 18 W/GRO-T8 lamps. The incident illumination was
measured outside the vessels with a Delta-T Logger (Delta-T Devices Ltd, UK),
equipped with a QS Quantum sensor.
Growth experiments were carried out on MultiDish 24 (Nunc, Denmark) cell
culture plates. They were performed at least in triplicate, each in 2 ml of BG11
media (Steiner et al, 1971). The starting inoculum depended on the maximum
cell count that could be reached by the specific species, in the range from 1 X 10s
to 5 X 10s cells I"1, and on the chlorophyll a content. The experiments were
carried out at three light intensities: in the dark, and at 4 and 40 u£ nr 2 s"1 at a
constant temperature of 20°C. The chronic influence of three different microcystin-RR concentrations of 1 X 10"7,5 X 10"7 and 1 X 10"* M on phytoplankton
in culture was tested. The microcystin tested was isolated in the laboratory and
its characteristics compared with a standard. Microcystin-RR was chosen as the
most frequent toxin in our waters, with a share up to >88% of total microcystin
contents (Sedmak and Kosi, 1997). The growth was determined by cell counting
in a Burker-Tiirk haemocytometer (Brand) under a microscope.
The permeability of cyanobacterial cells to microcystin-RR was tested in an
acute experiment. Microcystis aeruginosa Kuetz. cells, a non-toxic unicellular
strain, were concentrated by centrifugation to 107 cells ml"1 and exposed to the
toxin at a final concentration of 5.5 X 10~5 M MC-RR for 1 h at 0°C to prevent
cell division. The cells were washed twice with a 10-fold excess of cold BG11 to
remove the toxin from the medium. A cell growth experiment was then carried
out and proliferation determined as described previously.
Analytical methods
For chlorophyll a determination, the Vollenweider (1974) method was used. Two
techniques of cell harvesting were used: by centrifugation at 13 000 r.p.m. and
by filtering through glass microfibre Whatman GF/C filters (Whatman Ltd,
Maidstone, UK) and extracting with hot methanol.
694
Microcystins and bloom formation
Results
Phytoplankton diversity
The most striking similarity in all the water bodies studied was the sometimes
extreme reduction in species diversity, especially in blooms with higher microcystin contents with a rise of cyanobacterial species. In summer and autumn,
there are well over 30 different phytoplankton species (data not presented) in
individual water bodies, but during cyanobacterial blooms their count decreases
significantly (Tables I and II).
A correlation was found between the total microcystin content and the number
of phytoplankton species present in the blooms. Plotting the amount of microcystins against species diversity gave an exponential relationship. The data were
fitted by the exponential function:
y = A X exp(-fc X x) + Residue
The results in Figure 1 are arranged into two groups, where 10 ug microcystins
h 1 seems to be the boundary level. Above this concentration, the species diversity count falls to a minimal level.
Not only microcystin contents, but also bloom density, were negatively correlated with species diversity (Figure 2). The values for the locations with an operating outflow were dispersed, indicating unstable conditions during the bloom,
deriving from the hydraulic wash-out of cyanobacteria. In stable conditions, the
bloom could actually exert a much stronger influence on various species through
the resulting toxin release, and through the creation of adverse light conditions
in the water body. The influence of toxic cyanobacterial blooms on species diversity in stable conditions was clearly divided into two separate groups described
30-,
0
25-
pec iesd Iverslty (N° of
#
m
20-
; .
15-
•\
•\
10 -
4x10*/
Ctll numbtr
X
S
control
MC-RR 5x10"7 M
Fit statistics
SS = 260.8936
df= 12
r = 0.807477
#
3x10*-
2x10 5 -
\
1x10 5 -
5-
/
s
(0
0 0 0
10
20
30
40
50
60
Microcystins (ug/l bloom)
0
i
<
i
i
i
2
4
6
8
10
Days
Fig. L (A) The relationship between microcystin concentration and species diversity for all locations.
(B) The relationship between microcystin concentration and species diversity in small water bodies
with stable conditions during the bloom. The dashed vertical line indicates the 10 |ig microcystin I*1
bloom boundary.
695
RSedmak and G-Kosi
40-.
Q
0
•
O
s
*
30.
Locations with outflow
Locations without outflow < 10 (ig/l MCY
Locations without outflow > 10 |ig/l MCY
Intermediate value
o
%
i
20-
eCO
S
10-
0 - _
7
10*
10 "
10'
Bloom density (cells/1)
Fig. 2. The influence of bloom density on species diversity. The linear regression curves were calculated only for the locations without outflow, separately for groups with lower IMC and higher XMC
contents.
by linear regression curves with good correlation coefficients. We divided the
toxic blooms into two categories: blooms with low toxicity (<10 \ig total microcystins I"1) and blooms with high toxicity (>10 ug total microcystins I"1), as
suggested by the results presented in Figure IB.
Regression statistics
• locations
= <10 ug MC I"1 bloom
c
SS = 0.003893
d.f. = 3
52
= 0.001298
= 0.966158
r
r2 = 0.933461
• locations
1
c = >10 ug MC I" bloom
SS = 0.371687
(sum of squares)
(degrees of freedom)
df = 2
= 0.185843
r = 0.676206
r2 = 0.457254
52
(corr. data versus mo
The results from Figure 2 indicate that very high bloom densities (>1.8 x 108 cells
I"1) were associated with cumulative microcystin concentrations of >10 jig I"1
bloom.
Description of species
Cyanophyta. The most frequent bloom-forming genus in smaller water bodies is
Microcystis, with two species: M.aeruginosa and M.wesenbergii Kom. (Tables III
and IV). Other bloom-forming cyanobacteria are O.rubescens DC (Sedmak
and Kosi, 1991) and two (in our case) non-toxic filamentous representatives:
696
Microcystins and bloom formation
Anabaena flos-aquae (Lyngb.) Breb. and Aphanizomenon flos-aquae (Sedmak
and Kosi, 1997).
Other species present. In Lake Bled, only two green algae were found. Three
other species were abundant in this period: Dinobryon sociale Eh. (Chrysophyta),
a diatom belonging to the genus Stephanodiscus (Bacillariophyta) and Cryptomonas ovata Eh. (Pyrrhophyta). The non-toxic Aphanizomenon bloom in Ledava
reservoir had little effect on species diversity. The appearance of the Microcystis
bloom in Radehova reservoir decreased the number of species present, with only
one alga reaching a higher abundancy: Melosira varians Ag. (Bacillariophyta). A
heavy toxic bloom of the genus Microcystis in the Savci reservoir provoked a
drastic decrease in diversity. In 1994, there were only eight different species, three
of them belonging to the cyanobacteria. Koseze is a smaller water body with
blooms of Microcystis in both of the years 1994 and 1995. In 1994, the bloom
contained only very small amounts of toxins (Sedmak and Kosi, 1997). The
variety was affected, but there were two other species abundant: Uroglena americana Calk. (Chrysophyta) and Peridinium cinctum (O.F. Muller) Eh. (Pyrrhophyta). In the bloom of 1995, which was more toxic, but very weak in terms of
biomass, the diversity was high, with several species other than cyanobacterial
ones abundant. Among the gravel pits, the greatest diversity was found in the
weak bloom at Borovci. The sample from Lutverci contained only one diatom
species, M.granulata, with a small count. Both toxic Microcystis blooms in
Podgrad had a strong negative effect on diversity. In the bloom in 1994, an abundance of Ceratium hirudinella (O.F. Muller) Schrank (Pyrrhophyta) was
observed. In 1994, the effect on diversity was stronger, with only one species,
M.granulata (Bacillariophyta), being abundant besides cyanobacteria.
We inspected two locations, Podgrad and Savci, after 23 days in the phase of
bloom degradation. The cyanobacterial scum disappeared, although the species
composition remained similar. The overall species diversity increased in Podgrad
from 12 to 16 and in Savci from 8 to 11. In both locations, in the period of extensive blooming there were no Bacillariophyta species that reappeared with the
decline of cyanobacteria. An increase in Chlorophyta was especially evident in
Podgrad.
In vitro experiments
All algae tested exhibited an enhanced growth in the presence of microcystin in
the early stage and a resulting growth suppression afterwards, strongly dependent
on light intensity. A 10-fold increase in microcystin concentration did not result
in a proportional effect. On the contrary, the growth of M.aeruginosa was
enhanced with no inhibitory action even in the presence of higher microcystin
concentrations. In the absence of light, when the cells did not proliferate, we
detected no effects of microcystin on phytoplankton in culture.
Cyanobacteria. (i) Microcystis aeruginosa Kuetz. (non-toxic unicellular strain)—
chronic experiment. The influence of microcystin-RR on this most widely known
697
B-SedmaV and G-Kosi
2x10 7 -,
1.8x10 7
1.6x107 \
1.4x10' \
I
1x10'
c
S
8x10*:
6x10* i
4x10*2x10*-i
0-
2x10'
1.8x107
1.6x10T
1.4x10
B
B Control
• MC-RR 5x10"7 M
T
5
1.2X107
g
1x107
%
8x10*
6x10*
4x10*
2x10*
0J
8
10
12
14
16
18
Days
Fig. 3. The growth of M.aeruginosa at 40 \xE nr 2 s"1 illumination at two microcystin concentrations:
(A) XQr1 M and (B) 5 X 10"7 M microcystin-RR. The arrow indicates the onset of growth stimulation
(mean ± SE).
hepatotoxin producer depends strongly on light conditions and toxin concentrations. A higher toxin concentration shortened the response time and augmented
cell division (Figure 3).
The stimulation of cell division was evident at higher (5 X 10~7 M) microcystinRR concentrations, with a pronounced effect at lower light intensities (Figure 4).
(ii) Microcystis aeruginosa Kuetz. (non-toxic unicellular strain)—acute experiment. Cyanobacterial cells were exposed to microcystin-RR at low temperature
to prevent cell division. After 1 h incubation, the toxin was removed and the
cyanobacteria grown at 40 uE nr 2 s"1 illumination. Cell proliferation was
enhanced, indicating the permeability of the membrane to microcystin (Figure 5).
(iii) Chroococcus minutus (Kuetz.) Nag. The influence of microcystin-RR on
this cyanobacterium was strongly light dependent. Higher light intensities (40 uE
698
Microcystins and bloom formation
1x10' -,
• Control
• MC-RR 10"7 M
9x10*8x10*:
7x10* ^
I
|
c
%
O
6x10*:
4x10*3x10* :
2x10*:
1x10*:
0-
1x10 7 -,
9x10*:
8x10* •:
B
7x10* \
6x10*:
5x10*\
4x10* '3x10*2x10*1x10* '-
0
2
4
6
8
10
12
14
16
18
20
Days
Fig. 4. The growth of M.aeruginosa at 4 uE nr 2 s~' illumination at two microcystin concentrations:
(A) 10~7 M and (B) 5 x 10"7 M microcystin-RR. The arrow indicates the onset of growth stimulation.
m~2 s"1) caused growth suppression, with fluctuations in cell count suggesting cell
degradation. The onset of cell growth stimulation was positively correlated with
microcystin concentration (Figure 6). At a defined illumination and toxin concentration, stimulation of cell proliferation could be achieved (Figure 7A).
Algae
Coelastrum microporum Nag., Monoraphidium contortwn (Thur.) Kom.-Legn.
and Cryptomonas erosa Eh. The growth of C.microporum under low-light
conditions was strongly affected, resulting in inhibition of cell proliferation
(Figure 8).
The green alga M.contortwn, also growing successfully at low light intensities,
699
B£edmak and G-Kosi
2x10'
g Control
DMC-RR5.5x10"
5
M
1.5x10'
o
1x107
o
U
5x10*
0
2
4
6
8
10
12
Days
Fig. 5. Growth stimulation of M.aeruginosa after 1 h acute exposure to 5.5 X 10"5 M microcystin-RR
showed enhanced proliferation in the presence of microcystin-RR (Figure 9). The
same results were obtained with S.quadricauda (Chlorophyta) (data not
presented). In contrast, the initial stimulation of C.erosa (Cryptophyta) resulted
in cell loss at the same light and temperature conditions (Figure 10).
Phytoplankton species behave heterogeneously to a microcystin environment
under different light conditions. Unlike other phytoplankton species, cyanobacteria respond to a higher toxin concentration by shortening the response
time required for enhanced proliferation, indicating cell permeability to
microcystin-RR.
Discussion
From the field data, it is evident that small water bodies with stable conditions
better reflect all the noxious effects of toxic cyanobacterial blooms on the phytoplankton community in comparison to those with unstable conditions. Different
microcystins exhibit similar effects on mammalian cells, in vitro and in vivo (e.g.
Runnegar et al., 1993), or even on imaowsb a n d p i a n t s (MacKintosh et al., i§90).
We presume that thmy might a i s o exert a similar influence on phytoplankton
organisms at their cellular level. We therefore plotted cumulative values of the
microcystins present in a bloom againsi §pe§i§§ diversity. Water bodies with
higher total microcystin valugs In the bleeifl (5»10 H| MC H) have a higher negative impact on species diversity, and such a bloom fea"§&§§ higher cyanobacterial
cell densities (Figures 1 and 2). The calculated results offfllefoeystin contents per
cyanobacterial cell also confirm the last statement (Tables III and IV).
In vitro results demonstrate that microcystin-RR, particularly in low-light
conditions, can act heterogeneously on phytoplankton organisms. The effects can
range from augmentation of cell proliferation, as in the case of M.aeruginosa and
700
Microcystins and bloom formation
1.5x107 -,
• Control
Q MC-RR10"' M
1.2x10'
9x10 §
6x10*-
3x10*-
0 -I
0
2
4
6
8
10
12
14
16
Days
7
1.5x10 -,
B
• Control
• MC-RR 5x10"' M
1.2x10' -
9x10*:
%
O
6x10*-
3x10* \
,mT\
0
2
4
6
8
10
12
14
16
Days
Fig. 6. The growth of C.minutus at 40 nE in"2 s~' illumination at two microcystin concentrations:
(A) 10-7 M and (B) 5 x 10-' M microcystin-RR.
M.contortum (Figures 4 and 9), through inhibition, in the case of C.microporum
(Figure 8), or even result in cell loss, e.g. Cerosa (Figure 10). It seems that these
effects become evident only in specific light conditions, presumably depending on
the energetic light requirements of the species. These microcystin effects on
growth seem to be minor, but added together could represent an ecological
factor, promoting heavy bloom or scum formation. In order to act on cell proliferation, microcystins must penetrate the cell or at least the cell membrane. Microcystis aeruginosa cells are permeable to microcystin-RR, as demonstrated in the
acute experiment (Figure 5). Cells exposed to microcystin-RR proliferate faster
than controls. That means, in favourable conditions when the bloom is partially
formed and therefore turbidity rises, the microcystin producers proliferate faster
since the microcystin release in this phase is minimal (Watanabe et al, 1992). On
701
RSedmak and GJCosi
Table UL Toxic cyanobacterial blooms grouped according to total amount of hepatotoxins from
locations without stable conditions
Locations with outflow
Sampling
Year
site
Dominant species
Abundance
(cells 1-')
IMC
(Mgl-1)
SMC cell(Pg)
19.38
0.102
51.45
0.11
24.15
0.037
11.76
0.054
X 10'
1.88
0.024
x 10'
0.04
0.003
x 108
2.26
0.014
s
x 10
5.63
0.051
x 108
2.43
0.017
s
1.28
0.008
Blooms with high microcystin contents ( I M C > 10 ug I"1)
BOC
1996
M.wesenbergii 50%
1.9 X 108
M.aeruginosa 50%
HLP
1994
M.aeruginosa
4.7 X 10 s
HGD
1994
M.aeruginosa 95%
6.6 X 10 s
SAV»
1994
M.aeruginosa 92%
2.2 x 10 s
Blooms with low microcystin contents ( I M C < 10 ug I"1)
M.aeruginosa 90%
7.8
BOR
1994
M.wesenbergii 10%
M.aeruginosa
BOR
1996
1.5
1994
M.wesenbergii 75%
LUT
M.aeruginosa 25%
1.5
LUT
M.aeruginosa 70%
1996
M.wesenbergii 30%
1.1
M.wesenbergii 75%
POD b
1994
M.aeruginosa 25%
1.4
POD
M.wesenbergii 80%
1996
M.aeruginosa 20%
1.6
x 10
'Because of low rainfall, the outflow was inoperative,
intermediate value.
Table IV. Bloom composition and total hepatotoxin contents in locations subjected to wash-out
Locations with outflow
Sampling
Year
site
BUK
KOS
KOS
LED
SAV
1996
1994
1996
1995
1996
Dominant species
Abundance
(cells I"1)
IMC
(ug I'1)
IMC cell-'
(Pg)
M.wesenbergii
M.aeruginosa
M.aeruginosa 90%
Aph.flos-aquae
M.aeruginosa 70%
M.wesenbergii 30%
3.1
1.4
5.5
7.8
0
0.72
2.90
0
0
0.005
0.053
0
3.78
0.046
x 10'
x 10 s
x 10'
X 108
8.3 x 10'
the other hand, when microcystin-producing clones dominate the bloom, toxin
release could affect other species, leading to the disappearance of susceptible
ones.
On the basis of our results, we propose the following theory of bloom
formation. Eutrophic waters favour the growth of different phytoplankton
species that contribute to water turbidity. At the beginning, different cyanobacterial clones are evenly dispersed in the water body. Since the optimal light
intensities for toxin production are up to 40 uE nr 2 s"1, where the maximum
702
Microcystins and bloom fonnation
1.5x107 ,
I Control
O MC-RR10'7 M
1.2x10' -I
^
O
6x10* -I
3x10* J
0
2
4
6
8
10
12
14
16
Days
1.5x10' -,
B
1.2x107
• Control
D MC-RR 5 x 1 0 7 M
9x10*
E
O
6x10'
3x10*
0
2
4
6
8
10
12
14
16
Fig. 7. The growth of C.minutus at 4 uE m~2 s-' illumination at two microcystin concentrations:
(A) 10-7 M and (B) 5 x 10"7 M microcystin-RR.
toxicity and maximum ratio of toxin to protein are achieved (Utkilen and Gjolme,
1992), hepatotoxin producers are in a favourable position, due to water turbidity.
The buoyant cyanobacteria additionally generate even poorer light conditions,
allowing themselves to remain in a position to produce more toxins that are
released during the senescence of cyanobacteria (Watanabe et al, 1992; Lahti et
al, 1997). In combination with low light intensity, they can exclude the major part
of susceptible species from the bloom. In such a situation, microcystin production
is a self-enhancing mechanism augmenting cyanobacterial cell division, generating denser blooms than the non-producing cyanobacterial strains. We presume
that there is no environmental factor which converts an organism that does not
produce toxins into a toxin producer, but that toxin producers proliferate faster
and thus are able to capture the major share in the bloom.
703
RSedmak and CKosi
10'-,
g
10'
8x10*
5
6x10*
c
^
O
4x10*
2x10*
Fig. 8. Growth of C.microporwn at 4 \i£. m~2 s~' illumination at two microcystin concentrations:
(A) 10-7 M and (B) 5 X 10"7 M microcystin-RR.
Microcystins dissolved in water exert a chronic influence on other phytoplankton species. Although the concentrations are usually low, we cannot neglect
the microenvironment. This means the organisms that are in close proximity to
the toxin releaser may be subjected to essentially higher microcystin concentrations, influencing their growth (Figures 6, 7 and 8), which in unfavourable light
conditions may even lead to cell disintegration of certain species (Figure 10).
We have several field results confirming our hypothesis on the influence of
microcystins on species diversity. In Lake Bled, O.rubescens is a permanent
inhabitant together with sporadic blooms of An.flos-aquae and Aph.flos-aquae
(Vrhovsek et al, 1982, 1984; Sedmak and Kosi, 1991). During the mixed toxic
bloom of O.rubescens and An.flos-aquae in November 1994, we found an
extremely low number of planktonic species; only nine phytoplankton species
704
Microcystins and bloom formation
3x107 -
,_
e
2x10 7 -
.a
c
1x10'-
0J
0
2
4
6
8
10
12
14
16
Fig. 9. Growth stimulation of M.contortum; 1O"7 M microcystin-RR, illumination 4 jiE nr 2 s~'.
5x10*-
4x10*-
Z
3x1
O
2x10*-
I
°*"
1x10' -
0-
0
2
4
S
8
Days
10
12
Fig. 10. Cell loss of C.erosa due to the presence of microcystin. Light conditions: 4 jtE nr 2 s"1
were present, three of them belonging to the cyanobacteria. Because of the
ecophysical characteristics of the stratified lake, we can compare the effects of
cyanobacterial blooms in this type of lake to smaller water bodies. The actual area
in which the phytoplankton multiply is relatively narrow, comprising the metalimnion-hypolimnion layer. This means that the eventual toxin release from the
growth of Oscillatoria is also restricted to this area. Diatoms that are usually
present all the year around in the lake, such as Asterionella formosa Hass. and
Fragilaria crotonensis Kitt. (Vrhovsek et al., 1984, 1985; several unpublished
705
B^edmak and CKosi
results), disappeared completely during this bloom which contained very high
microcystin-YR levels (Sedmak and Kosi, 1997). Two other diatom species were
found in the samples: Stephanodiscus sp. in higher amounts and Cyclotella sp.,
demonstrating the availability of silicon. Asterionella formosa together with
F.crotonesis are described by several authors as very common and abundant
oligotrophic species present in association with different cyanobacteria
(Reynolds, 1980; Rosen, 1981; Bucka, 1987; Zhang and Prepas, 1996). However,
there are no data describing the presence of these two species in association with
hepatotoxic blooms.
In the Ledava reservoir, there was no bloom in 1994; nevertheless, in 1995 we
observed a non-toxic Aph.ftos-aquae bloom which did not markedly influence the
diversity of species composition. The Savci reservoir was in both years the site of
heavy Microcystis blooms. Owing to the low rainfall in 1994, the reservoir was
exposed to a mixed Microcystis bloom with high concentrations of microcystinRR, resulting in extremely low species composition (Table II). We noticed the
complete disappearance of Chrysophyta and Bacillariophyta species. In 1995,
because of the abundant rainfall and drainage through surface outflow of the less
toxic M.wesenbergii bloom, the phytoplankton community was less subject to
noxious effects. The result was an essentially higher species diversity. The gravel
pit in Borovci was the site of a weakly toxic Microcystis bloom with a more
heterogeneous composition, in spite of poor light conditions caused by water
turbidity arising from occasional gravel excavations. We noted the complete
absence of Chrysophyta and Euglenophyta taxa and several Bacillariophyta
species. Both the Lutverci and Podgrad fish ponds were dominated by similar
toxic mixed cyanobacterial blooms. In Podgrad, which is the main centre of a local
fishing community and therefore subjected to additional anthropogenic pollution,
we noted the disappearance of Chrysophyta and Bacillariophyta. Similarly, only
one diatom species was found in Lutverci. Two locations from the village of
Hotinja were added to our comparison and their detailed data will be discussed
elsewhere. In these two locations, we observed M.aeruginosa practically as a
monoculture, with an extremely low count of other species (Table I). These field
results show how entire higher taxonomic groups can temporarily be cleared from
the environment of a toxic cyanobacterial bloom, with several species being
inhibited or excluded.
Chlorophyta have a significant share in the phytoplankton species composition
of all the studied locations, with the exception of Lake Bled. Scenedesmus quadricauda is present in all highly eutrophic water bodies. Generally, the genera
Scenedesmus and Staurastrum are frequently associated with cyanobacterial
blooms (Reynolds, 1980; Rosen, 1981; Bucka, 1985; Lewis, 1986; Jensen et ai,
1994; Komarkova and Hejzlar, 1996). Inspection of the two locations after bloom
collapse revealed a rise in phytoplankton diversity with an increase in the abundancy of Chlorophyta species. It is well known that after the decline of cyanobacterial blooms, green algae became dominant (e.g. Lin, 1972). We ascribe the
intensified development of green algae after toxic cyanobacterial blooms to the
improvement of light conditions in the presence of microcystins, which can
augment their proliferation. The growth rates in culture of certain species of
706
Microcysrins and bloom formation
Scenedesmus can be doubled by adding water extracted from the end of a summer
bloom caused by M.aeruginosa (Hartman, 1960), which we presume contained
microcystins.
After bloom decomposition, large amounts of microcystins are dissolved in
water, with a simultaneous decrease in water turbidity. Toxins act as growth
stimulators for other species like S.quadricauda commonly present in cyanobacterial blooms and appearing in large numbers specifically after hepatotoxic
blooms (Bucka, 1989).
Acknowledgements
The authors wish to thank Andrej Blejec for his suggestions on statistical analysis, K.Stanic and I.Dragan for valuable assistance, and Professor C.S.Reynolds for
a fruitful discussion in Vigo. This research was supported by Ministry of Science
and Technology grant number L4-7403. Additional support was provided by the
Ministry of Agriculture and Forestry.
References
BenndorfJ. and Henning.M. (1989) Daphnia and toxic blooms of Microcystis aeruginosa in Bautzen
Reservoir (GDR). Int. Rev. Ges. Hydrobioi, 74,223-248.
Bucka.H. (1985) Ecology of some waters in the forest-agricultural basin of the river Brnica near the
Upper Silesian industrial region. Acta Hydrobioi., 27,493-508.
Bucka,H. (1987) Ecological aspects of the mass appearance of planktonic algae in dam reservoirs of
southern Poland. Acta Hydrobioi., 29,149-191.
Bucka.H- (1989) Ecology of selected planktonic algae causing water blooms. Acta Hydrobioi., 31,
207-258.
Codd.G.A. (1994) Blue-green algal toxins: water-borne hazards to health. In GoIding.A.M.B. and
Noah Nand Stanwell-Smith.R. (eds), Water and Public Health. Smith-Gordon, Great Britain, pp.
271-278.
Falconer,I.R., Burch,M.D., Steffensen.D.A., Choice,M. and Coverdale.O.R. (1994) Toxicity of the
blue-green alga (cyanobacterium) Microcystis aeruginosa in drinking water to growing pigs, as an
animal model for human injury and risk assessment. Environ. Toxicol. Water Quai, 9,131-139.
Harada,K-I., Matsuura,K., Suzuki.M., Oka,H., Watanabe,M.F., Oishi,S., Dahlem,A., Beasely.V.R.
and Carmichael.W.W. (1988) Chemical analysis of toxic peptides produced by cyanobacteria. /.
Chromatogr., 448, 275-238.
Harper J). (1992) Eutrophication of Freshwaters. Principles, Problems and Restoration. Chapman and
Hall, London.
Hartman,R.T. (1960) Algal metabolites of natural waters. In Tyron.C.A. and Hartman.R.T. (eds), The
Ecology of Algae. Special Publication No.2. Pymatuning Laboratory of Field Biology, University
of Pittsburg, pp. 38-55.
HindakJF. (ed.) (1978) Sladkovodne riasy. Slovenske pedagogicke nakladatelstvo, Bratislava.
JensenJ.P., JeppesenJE., Olrik.K. and Kristensen.P. (1994) Impact of nutrients and physical factors
to chlorophyte dominance in the shallow Danish lakes. Can. J. Fish. Aquat. Sci., 51, 1692-1699.
KomarekJ. (1958) Die taxonomische Revision der planktischen Blaualgen der Tschechoslowakei. In
Komarek j . and Ettl,H. (eds), Algologische Studien. Nakl. ESAV, Praha, pp. 10-206.
Komarek J. (1991) A review of water-bloom forming Microcystis species, with regard to populations
from Japan. Algol. Stud., 64,115-127.
KomarkovaJ. and HejzlarJ. (1996) Summer maxima of phytoplankton in the Rimov reservoir in
relation to hydrologic parameters and phosphorus loading. Arch. Hydrobioi., 136, 217-236.
Lahti.K., RapalaJ., Fardig,M, Niemela,M. and Sivonen.K. (1997) Persistence of cyanobacterial
hepatotoxin microcystin-LR in paniculate material and dissolved in lake water. Water Res., 31,
1005-1012.
707
aSedmak and GJCosi
Lewis.W.M.Jr (1986) Phytoplankton succession in Lake Valencia, Venezuela. Hydrobiologia, 138,
189-203.
Lin.C.G. (1972) Phytoplankton succession in an eutrophic lake with special reference to blue-green
algal blooms. Hydrobiologia, 39,321-334.
MacKintosh,C, BeattieJICA., Klumpp,S., Cohen,P. and Codd,G.A. (1990) Cyanobacterial microcystin-LR is a potent specific inhibitor of protein phosphatases 1 and 2A from both mammals and
higher plants. FEBS Lett., 264,187-192.
Mur,L.M., Gons,H-J- and Van Liere.L. (1978) Competition of the green alga Scenedesmus and the
blue-green alga Oscillatoria. Mitt Int. Ver. Limnol., 21,473-479.
Olson.T.A. (1964) Waterfowl tomorrow. In LindurskaJ.K. (ed.), Blue-greens. US Department of the
Interior Fish and Wildlife Service, Washington, DC, pp. 349-356.
Reynolds.C.S. (1980) Phytoplankton assemblages and their periodicity in stratifying lake system.
Holarctic Ecol., 3,141-159.
Rosen.G. (1981) Phytoplankton indicators and their relations to certain chemical and physical factors.
Limnologica (Berlin), 13,263-290.
Runnegar,M.T., Kong,S. and BerndtJV. (1993) Protein phosphatase inhibition and in vivo hepatotoxicity of microcystins. Am. J. Physiol., 256, G224-G230.
Sedmak.B. and Kosi.G. (1991) Algae and their toxins in national waters. Contemplations on cyanobacterial bloom, Aphanizomenonflos-aquae, in the Lake Bled. Vodoprivreda, 23,265-272 (English
Abstract).
Sedmak.B- and Kosi.G. (1997) Microcystins in Slovene freshwaters (Central Europe)—First report.
Nat. Toxins, 5, 64-73.
Sedmak,B., Kosi.G. and Kolar,B. (1994) Cyanobacteria and their relevance. Period. Bio/., 96,428-430.
ShapiroJ. (1990) Current beliefs regarding dominance by blue-greens: The case of the importance of
CO2 and pH. Verh. Int. Ver. Limnol., 24, 38-54.
Starmach.K. (1966) Cyanophyta-Sinice Glaucophyta-Glaucophity. Flora slodkowodna Polski, Tom 2,
Warszawa.
Starmach.K. (1980) Chrysophyceae-Zlotowiciowce. Flora slodkowodna Polski, Tom 5, Warszawa.
Steiner.R.Y., Kunisawa.R., Mandel,M. and Cohen-Bazire.G. (1971) Purification and properties of
unicellular blue-green algae (Order Chroococcales). BacterioL Rev., 35,171-205.
Utkilen.H. and Gjolme.N. (1992) Toxin production by Microcystis aeruginosa as a function of light in
continuous cultures and its ecological significance. Appl. Environ. Microbiol., 58,1321-1325.
Vollenweider.R.A. (1974) Primary Production in Aquatic Environments. Int. BioL Prog. Handbook
12. Blackwell Science, Oxford, 225 pp.
Vrhovsek,D., Kosi.G. and Zupan,M. (1982) The effect on water chemistry and phytoplankton of artificial inflow of the River Radovna into Lake Bled (Yugoslavia). Hydrobiologia, 96,225-242.
Vrhovsek,D., Kosi.G., Kralj,M. and Bricelj,M. (1984) The effect of three consequent sanation
measures in Lake Bled on physico-chemical and biological variables. Ekologija, 19,31-52.
VrhovSekJ)., Kosi.G., Kralj,M., BriceIj,M. and Zupan,M. (1985) The effect of lake restoration
measures on the physical, chemical and phytoplankton variables of Lake Bled. Hydrobiologia, 127,
219-228.
Watanabe,M.F., Tsuji.K., Watanabe.Y., Harada,K.-I. and SuzukiJU. (1992) Release of heptapeptide
toxin (microcystin) during the decomposition process of Microcystis aeruginosa. Nat. Toxins, 1,
48-53.
Zhang,Y. and Prepas^E.E. (1996) Regulation of the dominance of planktonic diatoms and cyanobacteria in four eutrophic hardwater lakes by nutrients, water column stability, and temperature. Can.
/. Aquat. Sci., 53, 621-633.
Received on September 2,1997; accepted on November 21,1997
708