ACTA BIOLOGICA SLOVENICA LJUBLJANA 2002 Vol. 45, Št. 1: 17 - 30 Harmful cyanobacterial blooms in Slovenia - Bloom types and microcystin producers Škodljiva cianobakterijska cvetenja v Sloveniji - Tipi cvetenj in proizvajalci mikrocistinov Bojan SEDMAK and Gorazd KOSI National Institute of Biology, Večna pot 111 , 1001 Ljubljana, Slovenia E-mail: bojan.sedmak@uni-lj.si Abstract. Up to now, research on cyanobacteria and their biologically active substances has been directed principally towards their harmful effects on humans, and little has been done to elucidate their ecological role. In order to understand better the biological success of cyanobacterial blooms, and in order to be able to compare the results of different scientific investigations, we must find and agree ona definition of the phenomenon. We propose a definition ofharmful cyanobacterial blooms based on the OECD boundary system of eutrophication with the addition of phycocyanin values. We have found a direct linkage between the trophic conditions in the water-bodies and the frequency of formati on of cyanobacterial blooms. Specific toxic species and their strains have been studied intensively. However, in order to elucidate the mechanisms that enable cyanobacteria to overtake eutrophic water bodies we must change our approach. Cyanobacterial blooms should not be treated merely as different species or strains but as superorganisms. It is their intraspecific diversity that permits cyanobacteria to be successful in a variable water environment. We here focus attention on microcystin producers and microcystins as an adaptation to the limited light conditions, which arise in cyanobacterial blooms. The conclusions are illustrated with some HC ··-··-.. ·~ L ....... L __________ _ --------···-··· ................... _ .....•... '················;.............................. ULTRAOLIGOTROPHIC . ~ ........................ .. , ............ , ........... . , ..... . ·-·-·-·-·-·-·- ·-·-·-·-·-·-·-·-·-·-·-·-·-" l~ Legend: OECDValue Hypertrophic (Chlorophyll mglm') Cyanobacterial b/ooms -·-·-·-·-·-L Eutrophic Mesotrophic Seeming1y oligotrophic Oligotrophic Ultraoligotrophic 225 8-25 2.5 • 8 š2.5 š2.5 š1 Frequent and regular Frequent and regular Occasional Frequent and regular metalimnetic No No Figure!: Multivariate cluster analysis using a modified Bray-Curtis method (CLARKE & WARWICK 1990). Comparison of phytoplankton associations in surface water bodies of different trophic categories in Slovenia, based on chlorophyll contents of the water. Slika 1: Multivariantna klasterska analiza po prilagojeni Bray-Curtisovi metodi (CLARKE & W ARWICK 1990). Primerjava fitoplanktonskih združb v slovenskih površinskih vodnih telesih različnih trofičnih kategorij uvrščenih na podlagi vsebnosti klorofila. 22 Acta Biologica Slovenica, 45 (!), 2002 Cyanobacterial bloom types in Slovenia There are three basic types of cyanobacterial blooms in Slovenia which differ in the origin of nutrients: a.) Planktonic blooms build up in eutrophic and hypereutrophic water bodies with nutrients evenly dispersed in the water, and where nutrient availability is influenced by diumal stratification (ali bloom forming species involved). b.) Metalimnetic blooms build up in deeper mesotrophic and eutrophic reservoirs and in "seemingly" oligotrophic lakes where the nutrients become available asa consequence of seasonal stratification (predominantly filamentous species) . c.) Benthic blooms (cyanobacterial mats) build up in eutrophic and mezotrophic shallow water bodies, where the benthic cyanobacteria utilize nutrients from the sediment (Oscillatoria princeps). In a temperate climate with seasonal changes, as in Slovenia, we are faced with stratification in ali water bodies when the vertical mixing is weak. Therefore regard must also be paid to nutrient availability asa factor that triggers the beginning of cyanobacterial blooms (REYNOLDS 1984a). However, once the cyanobacterial bloom has started to form, light takes over as the major limiting factor in phytoplankton growth. Basic cyanobacterial freshwater bloom configurations Cyanobacterial blooms appear in different forms, depending on cyanobacterial abundance and on climatic and meteorological conditions. - Dispersed blooms occur at the beginning of bloom formati on and can appear secondarily as the consequence of vertical mixing due to high wind velocities (GEORGE & EDWARDS 1976). In such an environment the lower light conditions are mainly due to mutual shading of the plankton. - Metalimnetic blooms occur in clear stratified lakes, where light penetrates beyond the depth of the epilimnion. Such blooms ari se where opposing gradients of irradiance and nutrients are established due to the mobilisation of nutrients from the lake bottom (GANF & OLIVER 1982, KoNOPKA 1989). In upper layers the light conditions are good and the position of cyanobacteria is due to their regulation of buoyancy (REYNOLDS & WALSBY 1975), - Surface blooms occur in calm weather and good insolation, when the speed of the wind is less than 2 - 3 m s·1, resulting in low mixing rates (WEBSTER & HuTCHINSON 1994 ). The light conditions below the bloom are bad, butare rescued by buoyancy regulation within the cyanobacterial population. Cyanobacteria altematively migrate towards the surface in a constant exchange of celi s and colonies at the water surface. Under favourable conditions such surface blooms may give the mistaken impression of a persistent bloom (KROMKAMP & W ALSBY 1990). - Scums originale from persistent blooms where there is a physical restraint on vertical movement (IBELINGS & MuR 1992, W ALSBY 1994). Cyanobacterial scums prevent the penetration of light to deeper layers . The cells at the surface are often severly damaged by the high light intensities that can induce dehydration and celi senescence. The most evident common characteristic of cyanobacterial blooms is the light environment, which ranges from low to very low and in extreme circumstances to even almost complete darkness. In severa! publications Mur emphasises the importance of light for cyanobacterial dominance. Results from competition experiments on the growth of Scenedesmus and Oscillatoria have shown that B. Sedmak, G. Kosi: Harmful cyanobaterial blooms in Slovenia - Bloom types and microcystin ... 23 cyanobacteria can reach higher growth rates than green algae only under extreme light limitation (e.g. MuR 1983). Light limitation can be provoked by dispersed particles of different origin, or by mutual shading of phytoplankton. In shallow eutrophic water bodies there are different periodic progressions from one dominant phytoplankton assemblage to another (REYNOLDS 1980). These environments reduce the light to such an extent that only cyanobacteria remain competitive, giving rise in tirne to a massi ve population and out-competing other autotrophs. With the appearance of cyanobacteria the light availability rapidly decreases. Gas vesicles present in buoyant cyanobacterial species induce additional horizontal light scattering that diminishes further light availability in deeper layers (WALSBY 1994). Cyanobacterial blooms and microcystins in Slovenia Our interpretations are based on results obtained from natura) populations in Slovene water bodies, and on laboratory experiments with isolated cyanobacterial species and strains grown in vitro under controlled conditions. Our most frequent bloom forming cyanobacterial genera are Microcystis, Anabaena, Aphanizamenon and Oscillatoria. Dispersed blooms are planktonic blooms, occuning in ali eutrophic water bodies at the beginning of bloom formation or asa product of vertical mixing. They can be either toxic or non-toxic. The strains of cyanobacteria that are present at the beginning of bloom formation are in the majority of cases of non microcystin producing types. lsolates of strains from different natura( blooms have led to the identification of severa! non- producing and diverse producing strains of cyanobacteria. The isolation of a relatively high percentage of non-producing strains from toxic natura! blooms can be explained by the fact that, for a successful isolation, single cells or filaments or small colonies are used. It has been namely demonstrated that larger colonies in most cases belong to producing strains (JuNGMANN & al. 1996). This observation leads to the assumption that producing strains are better adapted to bloom conditions, since they proliferate with faster dividing rates than non-producing strains, resulting in bigger colonies. Of course there are also severa) producing and non-producing strains that do not aggregate in colonies and proliferate in single celi configuration. The evolution of a cyanobacterial bloom is a highly dynamic process in which a broad variety of strains are involved. The constant changes in the light environment in the bloom, due to the growth of cyanobacteria on the one hand and to meteorological and hydrological changes on the other, give different strains the opportunity to proliferate. With the aggravation of light conditions, strains that are better adapted predominate. It has long been known that Microcystis is non-toxic at the beginning of the growing season, but develops high toxicity during the first strong biomass increase (BENNDORF & HENNING 1989). Already in the sixties it was established that the possibility of a cyanobacterial bloom being toxic is over 50% (OLSON 1964 ). In our investigations this rises to o ver 80% (SEDMAK & al. 1994, SEDMAK & Kos, 1997a). The main difficulty in comparing such results from the literature lies in the poor definition of the bloom. Adopting our definition for cyanobacterial blooms and taking into consideration only the principal cosmopolite microcystin-producing genus Microcystis, the statement that cyanobacterial blooms evolve to toxic becomes a rule. We conclude that producing strains that prevail overwhelmingly in the bloom are better adapted to a low Iight environment. When we concentrate such dispersed blooms we can detect diverse microcystins that could originate also from different strains. 24 Acta Biologica Slovenica, 45 ( 1), 2002 Ali filamentous and non-filamentous bloom-forming genera appear occasionally in the form of dispersed blooms. Metalimnetic blooms are common in deeper stratified lakes and reservoirs . The main bloom forming species in Slovenia are filamentous Aphanizomenonflos-aquae, Anabaenaflos-aquae and Oscillatoria rubescens (SEDMAK & Kosi 1997a, SEDMAK & Kosi 1991). They may be either producing or non-producing. We are concemed with Lake Bled, since it is the main centre of tourisrn in the region . The term "seemingly" oligotrophic is used because the productivity ofthelake is norrnally low and the inflows are rich and permanent. The nutrients in the phase of summer stratification diffuse from the lake bottom and support metalimnetic blooms. O. rubescens grows alrnost every year when water stratification is established. In favourable meteorological and climatic conditions Oscillatoria rnigrates to the surface forming a surface bloorn or even scurn frequently covering alrnost the entire lake surface. Such blooms can persist on the surface even in January and can grow underthe ice cover. In such cases we can normally detect microcystin-YR in bloom sarnples. It appears that the strain capable of MC-YR production is the best adapted to counter the lake environrnent. Lake Bled has two marked depressions which function as two independent sites of cyanobacterial growth. For this reason, unusual surface blooms can be observed as separate blooms of O. rubescens and An. flos-aquae, which subsequently merge in a unique mixed surface bloom. Blooms composed of equal parts of O. rubescens and M. aeruginosa have also been observed. Surface blooms and scums prevail in summer and autumn in smaller eutrophic and hypertrophic water bodies such as reservoirs, fishponds and abandoned gravel pits. The rnain species are Microcystis aeruginosa, Microcystis wesenbergii andAn.flos-aquae (Sedmak & Kosi 1997b). Microcystis species are almost always toxic. Occasionally there are also blooms of Aphanizomenonflos-aquae, Oscillatoria limnetica and Oscillatoria agardhii, which till now were ali identified as non-producing. Mixed blooms are also common. M. aeruginosa appears frequently together with M. wesenbergii or An.flos- aquae. Benthic cyanobacterial species like O. princeps can also rise to the surface and form surface blooms in the case ofhigh proliferation rates. So far we have not been able to detect productive strains of this species. However there are some cases when rnicrocystins have been produced by benthic cyanobacteria (MEZ. et al. 1997). Ali blooms have at least two things in common, high fluctuations in oxygen content and high light limitation. Additionally, the Microcystis blooms and scums are, to a large extent, associated with microcystin production. These bloorns in a similar environment in North-eastem Slovenia ali end with an almost identical pattem of production of microcystins-RR and-LR (SEDMAK & Kosi 1977a). Again we can assume that these productive strains are the best adapted to take over the highly eutrophic water bodies in the specific environment of the region. A concise survey of microcystin production and their possible role Today it is perfectly clear that the ability of a strain to produce microcystins depends on its possession of the necessary genes (MEIBNER et al.1996; DnTMANN et al. 1997). Thus there are strains able to produce microcystins to different extents and others that are not able to produce them at ali . Meanwhile, the amount of production is dependent principally on ecological conditions ( e.g. W ATANABE & 01stt1 1985). So it is obvious that the success of a particular strain depends on its adequacy under given conditions. There is great intraspecific biodiversity in the production of those biologically active substances that benefit the producing organisms, which is also the case in microcystin synthesis (NEILAN et al. 1999). So in the space of tirne from the origin of a bloom to its senescence, we have a B. Sedmak, G. Kosi: Harmful cyanobaterial blooms in Slovenia - Bloom types and microcystin ... 25 series of physiologically diverse cyanobacteria that can be successful to different degrees in various conditions, even in the framework of the same species. The possibility of natura! genetic transformation in bloom conditions is very low due to the short tirne span of the bloom and relatively low dividing rates of cyanobacteria. So far it seems that there is no single factor responsible for the variation in toxicity of cyanobacteria. Fram various data in vitro as well as from data obtained from natura! blooms, it is evident that the differences in microcystin production between strains are very large, microcystin contents ranging from zero to 1.5% of cyanobacterial biomass according to JvNGMANN and co-workers (] 996) and 1.84% microcystins /dry weight according to UTKILEN and GJ0LME (1992) . In our analyses total microcystin content can reach 2% of cyanobacterial dry weight (calculated value 0.64 pg/cell) in natura! populations (SEDMAK & Kosi 1997a). On the other hand, the chlorophyll a content of M. aeruginosa is on average 1.5% of celi dry weight (0.26 - 0.43 pg/cell) (REYNOLDS 1984b ). Such huge microcystin production, comparable to the content of the indispensable chlorophyll a, emphasizes the ecological importance of these substances for the producing cyanobacteria. The most recent findings indicate that microcystin synthesis proceeds via a multienzyme complex consisting of both peptide synthetase and polyketide modules (KAEBERNICK & NEILAN 2001 ). In such a synthesis significant cellular energy is required. The role of microcystin should be correspondingly great. As already mentioned, cyanobacteria were among the first organisms to inhabit our planet. It is more probable that their evolution involved biologically active substances that would aid them in their adaptive capabilities rather than harm rivals that did not even exist. Fram a quick survey of scientific data we can summarize that there is no evolutionarily adaptive value for cyanobacteria to kili land animals and fish (JuNGMANN et al. 1996). It can be expected that cyanobacteria and their products internet primarily with organisms in the same environment. Microcystins are not likely defence substances against grazers, since experiments have demonstrated that they are toxic to zooplankton only at very high concentrations (DE MoTT et al. 1991) and that there are other, more effective substances against them isolated from cyanobacteria (e.g. N1zAN et al. 1986). Therefore according to Occam's razar (WILLIAM OCKHAM 1285 - 1349) whereby »unnecessary assumptions should be abandoned«, the assumption that microcystins are defence substances is, in our opinion, unnecessary. We have proposed that there is no environmental factor capable of converting a non-producing strain into a producing strain (SEDMAK & Kosi 1998a, SEDMAK & Kosi 1998b ), but that there must be an ecological factor that augments the growth of microcystin producing cyanobacteria in order to prevail over other species and strains. In the last decade, research has been focused on environmental factors that could trigger microcystin production, rather than looking at cyanobacterial species as complex mixtures of different strains with diverse adaptive values. Light availability and microcystin production Work on microcystin production has been concentrated on the study of specific producing strains of cyanobacteria, in order to find optimal conditions where they proliferate and produce and with the aim of understanding how they would behave in bloom conditions. However, in specific bloom conditions, they may represent changeable proportion ofthe cyanobacterial biomass . In nature it is the environment that gives the opportunity to the fittest. During the evolution of the bloom there is a constant change of conditions. Specific strains are favoured and, with their proliferation, there is, in tum, an additional change in environmental conditions that offers the opportunity to another better adapted strain to propagate. With the growth of the bloom, light conditions become worse and only strains capable of proliferation under such extreme environments can prevail. In our opinion, microcystin producing strains are better adapted to low light conditions than non-producing strains (SEDMAK 26 Acta Biologica Slovenica, 45 (1), 2002 2001 ). This explains why the blooms become more toxic with the increase in cyanobacterial biomass. It also explains the poor permeability of cells to microcystins, which ate primarily designed to influence their own physiology. 0RR and JoNES ( 1998) have shown that the highest microcystin concentrations are produced under conditions optimal for celi growth. There is a high probability therefore that under optimal conditions there will be an optimal production of ali the microcystin variants that a strain is capable of producing. These optimal conditions for producing cyanobacteria coincide with relatively low light conditions. It is necessary therefore to be able to estimate the availability of light in eutrophic water bodies with abundant phytoplankton growth. The optimal light intensities for producing M. aeruginosa strains have been estimated as being less than 40 microeinsteins m·2 s·1 (UTKILEN & GJ0LME 1992). Similar values have been reported for other microcystin producing cyanobacteria (RAPALA et al. 1997). Such intensities are normally found in cyanobacterial blooms at a depth of about 1 m (UTKILEN & GJ0LME 1992). From these data it is evident that producing cyanobacteria ate adapted to the low light conditions characteristic of an already existing bloom. Cyanobacteria that start such an environment usually belong to the non-producing or poorly producing strains. Microcystin production is energy consuming and becomes an advantage only in an adequate environment. We have found a strong positive correlation between microcystin production and cyanobacterial cell concentration in the bloom, suggesting that producing strains ate capable of more dense bloom and scum formati on and of survival in consequent low light environment (Fig 2). 250 200 • 150 "' o :1: o, :i. R= 0.846695 100 • • 50 • o...----~- --~---~ 0,00E+00 4,00E+0S 8,00E+05 1,20E+06 Cyanobacterial cell concentration (cells/1) Figure 2: The relationship the cumulative values of produced microcystins (IMC) in the bloom and the cyanobacterial celi concentration. Slika 2: Razmerje med vsebnostjo vseh mikrocistinov (IMC) in koncentracijo cijanobakterijskih celic v cvetu. There have been a few attempts to find a linkage between photosynthesis and microcystin production. Microcystins have been found primarily in the thylakoid and nucleoid regions (SHI et al. 1995) and it is believed that the ADDA moiety of the toxin may bind to the thylakoid. This association with the photosynthetic appatatus of cyanobacterial cells may indicate a function in the light harvesting and chromatic adaptation mechanisms exhibited by these organisms (ORR & JoNES 1998) Light dependent processes are essential to both prokaryotic cyanobacteria and eukatyotic algae, which compete in the same habitat. Thus the production of biologically active substances that gives an advantage in this crucial area is clearly supported in the process of evolution. B. Sedmak, G. Kosi: Harmful cyanobaterial blooms in Slovenia - Bloom types and microcystin ... 27 Can microcystins influence other phytoplankton species in the environment? Despite the established opinion that microcystins are generally not cell-permeable, except the hepatocytes, which have a specific uptake system, we have demonstrated that they can influence the growth of different phytoplankton species in cul ture even at low concentrations ( 10-1 M) (SEDMAK & Kosi 1998a, SEDMAK & Kosr 1998b). Recent investigations have confirmed the possibility of non- specific translocation of oligopeptides (to undecamer) across plasma membranes (OELHKE et al. 1997). Microcystins are heptapeptides and as such they can pass the celi membrane. Whether the release of microcystins is due to celi death or to celi leakiness, it remains a fact that microcystins can be detected in the environment during celi proliferation (RAPALA et al. 1997). In such a way, non-producing strains also may persist longer in the bloom, preserving a bigger biodiversity of strains. In vitro experiments have namely confirmed that non-producing strains exposed to microcystins achieve higher proliferation rates under low light conditions than otherwise (SEDM AK & Kosr 1998a, SEDMAK & Kosr 1998b). Such diverse cyanobacterial association can