Dominance of Cyanobacteria (Blue-Green Algae) in Lakes and Reservoirs in Serbia: Can it be Predicted in Time to React with Management Measures?

Prvoslav Marjanović1, Dušan Kostić1, Dragica Vulić1, Marko Marjanović1

 

1 Jaroslav Černi Institute for the Development of Water Resources, Jaroslava Černog 80, 11226 Pinosava-Belgrade, Serbia; E-mail: This e-mail address is being protected from spambots. You need JavaScript enabled to view it

 

Abstract

The idea to use a TN:TP ratio as the main predictor of the risk of occurrence and blooms of cyanobacteria in lakes and reservoirs in Serbia, comes from the suggestion made by Smith (1983), that low TN:TP ratios promote dominance of cyanobacteria. We have analyzed data on abundance and biomas of algae (including % biomass of cyanobacteria, and concentration of total nitrogen and total phosphorous for 31 lakes for the period from 2004 to 2015. The number of samples taken in each lake in this period varied between the lakes and ranged from 56 to 252 samples. The available data was used to calculate the average annual water column values for parameters of interest for each of the 27 lakes/reservoirs. (% Cyanobacteria and TN:TP ratio). Our findings suggest the following: Dominance and blooms of cyanobacteria in lakes and reservoirs in Serbia do not occur if the mean annual water column TN:TP ratio is less than 29. This is in complete agreement with the findings of Smith (1983). It is extremely interesting that such an agreement seems to hold even for multi seasonal analyses of, not only epilimnetic concentrations, but whole water column concentrations suggesting that the "memory" and "history" of cyanobacterial occurrence play a significant role in the development of cyanobacterial blooms in the lakes and reservoirs in Serbia. Furthermore, blooms of cyanobacteria in the lakes/reservoirs in Serbia tend to occur only if cyanobacteria were present in the lake/reservoir during the previous growing season in excess of 20 % by weight. The data also suggests that the development of a cyanobacterial blooms is a multi-seasonal phenomenon in most lakes/reservoirs in Serbia. Risk assessment of cyanobacterial blooms and interference with the water supply systems can therefore be associated with the whole water column TN:TP ratio and presence of cyanobacteria in the previous period (growing season or quarter) in general. Our findings and the risk assessment framework presented alone are not sufficient to explain the cause and effect chain of events with respect to the occurrence or absence of cyanobacteria in all lakes and reservoirs in Serbia. Our data nevertheless suggests that lakes having whole water column TN:TP ratios > 29 will typically show very low proportions of cyanobacteria. The practical implications of these findings are that modification of TN:TP ratios can be a viable strategy for controlling the risk of cyanobacterial blooms. The methods and means for controlling TN:TP ratios need further study and elaboration.

Keywords: cyanobacteria, water supply, risk assessment, TN, TP, ratio.

Introduction

"A pet child has many names". This proverb is applicable to the topic of the paper which looks at blue-greens, blue-green algae, myxophyceaens, cyanophyceans, cyanophytes, cyanobacteria, cyanoprokaryotes, which are all names referring to the same group of organisms. The proliferation of names highlights the importance of these organisms in the development of biology as a science. From their earliest observation and recognition by botanists (Geitler, 1932, Linné, 1753, Vaucher, 1803 ), and onwards to their depiction in literature (Anagnostidis and Komárek, 1985; Staley et al., 1989) and papers, coupled with the amazing combination of properties found in algae and bacteria which these organisms exhibit, have been a source of fascination and attraction for many scientists. Cyanobacteria also provide an extraordinarily wide-ranging contribution to human affairs in everyday life (Ingrid Chorus and Jamie Bartram, 1999) and are of economic importance in many cases.

Both the beneficial and detrimental features of cyanobacteria are of considerable significance from a human perspective. Their extensive growth can create a considerable nuisance for management of inland waters (water supply, recreation, fishing, etc.) and they also release substances into the water which may be unpleasant (Jüttner, 1987) or toxic (Gorham and Carmichael, 1988). Water quality problems caused by dense populations of cyanobacteria are intricate, many and various (Skulberg, 1996b) and can have great health and economic impacts.

Concern about the effects of cyanobacteria on human health has grown in many countries in recent years for a variety of reasons. The Working Group on Protection and Control of Drinking-Water Quality (WHO) identified cyanobacteria as one of the most urgent areas in which guidance is required. During the development of the WHO Guidelines for Safe Recreational Water Environments, it also became clear that health concerns related to cyanobacteria should be considered, especially in fresh water reservoirs used for water supply and have become a subject of increasing public and professional interest. Abundant growth of cyanobacteria in water reservoirs creates severe practical water supply problems. The development of strains containing toxins is a common experience in polluted inland water systems all over the world.

When temperature, light and nutrient status conditions are right, surface waters can host increased growth of algae or cyanobacteria. Where such proliferation is dominated by a single (or a few) species, the phenomenon is referred to as an algal or cyanobacterial bloom.

Surveys on cyanobacteria and cyanotoxins have been primarily ecological and biogeographical. Early surveys in a number of countries, including Australia, Canada, Finland, Norway, South Africa, Sweden, the UK and the USA involved toxicity testing by mouse bioassay. Surveys during the 1990s employed more sensitive and definitive methods for the characterization of the toxins, such as chromatographic or immunological methods. (Chorus and Bartram, 1999). These studies provide more superior methods for estimating the expected range of concentrations in a given water body and season.

In regions using surface waters affected with cyanobacteria as a source for drinking water, actual toxin exposure will depend strongly on the method of water abstraction and treatment. In comparing the available indications of hazards from cyanotoxins with other water related health hazards, it is conspicuous that cyanotoxins have caused numerous fatal poisonings of livestock and wildlife, but no human fatalities due to oral uptake have been documented. Human deaths have only been observed as a consequence of intravenous exposure through renal dialysis. (Chorus and Bartram, 1999). Cyanotoxins are rarely likely to be ingested by humans in sufficient amounts for an acute lethal dose. Thus, cyanobacteria are less of a health hazard than pathogens such as Vibrio cholerae or Salmonella typhi. (Chorus and Bartram, 1999).

The combination of available knowledge on chronic toxicity mechanisms (such as cumulative liver damage and tumor promotion by microcystins) with that on ambient concentrations occurring under some environmental conditions, shows that chronic human injury from some cyanotoxins is likely, particularly if exposure is frequent or prolonged at high concentrations.

All this points to the need to predict the occurrence and blooms of cyanobacteria in water supply reservoirs and access the risk of exposure in order to employ control measures and prevent human exposure to cyanobacteria and their metabolites via water.

The ability to predict the occurrence and dominance of cyanobacteria is primarily a function of the ecology of cyanobacteria and factors controlling their growth rates.

 

Ecological Factors of Importance

Cyanobacteria have a number of special properties which determine their relative importance in phytoplankton communities. However, the behavior of different cyanobacterial taxa in nature is not homogeneous because their eco physiological properties differ. An understanding of their response to environmental factors is fundamental for setting water management objectives. Because some cyanobacteria show similar ecological and eco physiological characteristics, they can be grouped by their behavior in planktonic ecosystems as "ecostrategists", typically inhabiting different niches of aquatic ecosystems. (Chorus and Bartram, 1999)

The majority of cyanobacteria are aerobic photoautotrophs. Their life processes require only water, carbon dioxide, inorganic substances and light. Photosynthesis is their principal mode of energy metabolism. In the natural environment, however, it is known that some species are able to survive long periods in complete darkness. Furthermore, certain cyanobacteria show a distinct ability for heterotrophic nutrition (Fay, 1965). They flourish in water that is salty, brackish or fresh, in cold and hot springs, and in environments where no other microalgae can exist.

Freshwaters with diverse trophic states are the prominent habitats for cyanobacteria. Numerous species characteristically inhabit, and can occasionally dominate, both near-surface epilimnetic and deep, euphotic, hypolimnetic waters of lakes (Whitton, 1973). Another remarkable feature is their ability to survive extremely high and low temperatures. Cyanobacteria are inhabitants of hot springs (Castenholz, 1973), mountain streams (Kann, 1988), Arctic and Antarctic lakes (Skulberg, 1996a) and snow and ice (Kol, 1968; Laamanen, 1996). The cyanobacteria also include species that run through the entire range of water types, from polysaprobic zones to katharobic waters (Van Landingham, 1982).

Like algae, cyanobacteria contain chlorophyll a as a major pigment for harvesting light and conducting photosynthesis. They also contain other pigments such as the phycobiliproteins which include allophycocyanin (blue), phycocyanin (blue) and sometimes phycoerythrine (red) (Cohen-Bazire and Bryant, 1982). These pigments harvest light in the green, yellow and orange part of the spectrum (500-650 nm) which is hardly used by other phytoplankton species. Many cyanobacteria are sensitive to prolonged periods of high light intensities. Long exposures at light intensities of 320 μE m-2 s-1 are lethal for many species (Van Liere and Mur, 1980). However, if exposed intermittently to this high light intensity, cyanobacteria grow at their approximate maximal rate (Loogman, 1982). This light intensity amounts to less than half of the light intensity at the surface of a lake, which can reach 700-1,000 μE m-2 s-1.

Cyanobacteria are further characterized by a favorable energy balance. Their maintenance constant is low, they require little energy to maintain cell function and structure (Gons, 1977; Van Liere et al., 1979). As a result of this, the cyanobacteria can maintain a relatively higher growth rate than other phytoplankton organisms when light intensities are low. The cyanobacteria will therefore have a competitive advantage in lakes which are turbid due to dense growths of other phytoplankton. This was demonstrated in an investigation measuring growth of different species of phytoplankton at various depths in a eutrophic Norwegian lake (Källqvist, 1981). The results showed that the diatoms Asterionella, Diatoma and Synedra grew faster than the cyanobacterium Planktothrix at a 1 m depth, while the growth rate was about the same for all these organisms at a 2 m depth. At the very low light intensities below 3 m only Planktothrix grew. The ability of cyanobacteria to grow at low light intensities and to harvest certain light-specific qualities enables them to grow in the "shadow" of other phytoplankton. Van Liere and Mur (1979) demonstrated competition between cyanobacteria and other phytoplankton. Whereas the green alga (Scenedesmus protuberans) grew faster at high light intensities, growth of the cyanobacterium (Planktothrix agardhii) was faster at low light intensities Therefore, in waters with high turbidity they have better chances of out-competing other species. This can explain why cyanobacteria, which can grow under very poor nutritional conditions, often develop blooms in nutrient-rich eutrophic waters.

Many planktonic cyanobacteria contain gas vacuoles (Walsby, 1981). These structures are aggregates of gas-filled vesicles, which are hollow chambers with a hydrophilic outer surface and a hydrophobic inner surface. A gas vesicle has a density of about one tenth that of water (Walsby, 1987) and thus gas vesicles can give cyanobacterial cells a lower density than water.

Slow growth rates require long water retention times to enable a bloom of cyanobacteria to form. Therefore cyanobacteria do not bloom in water with short retention times. A. comprehensive overview of mechanisms determining the growth rates of planktonic algae and cyanobacteria under different field conditions is available in Reynolds (1987).

Because cyanobacterial blooms often develop in eutrophic lakes, it was originally assumed that they required high phosphorus and nitrogen concentrations. This assumption was maintained even though cyanobacterial blooms often occurred when concentrations of dissolved phosphate were lowest.

Experimental data have shown that the affinity of many cyanobacteria for nitrogen or phosphorus is higher than for many other photosynthetic organisms. This means that they can out-compete other phytoplankton organisms under conditions of phosphorus or nitrogen limitation. In addition to their high nutrient affinity, cyanobacteria have a substantial storage capacity for phosphorus. They can store enough phosphorus to perform two to four cell divisions, which corresponds to a 4-32 fold increase in biomass. However, if total phosphate rather than only dissolved phosphate is considered, high concentrations indirectly support cyanobacteria because they provide a high carrying capacity for phytoplankton. High phytoplankton density leads to high turbidity and low light availability, and cyanobacteria are the group of phytoplankton organisms which can grow best under these conditions. (Ingrid Chorus and Jamie Bartram, 1999)

A low ratio between nitrogen and phosphorus concentrations may favor the development of cyanobacterial blooms. A comparison between optimum N:P ratios for eukaryotic algae (16-23 molecules N:1 molecule of P) and optimum rates for bloom-forming cyanobacteria (10-16 molecules N: 1 molecule P), shows that the ratio is lower for cyanobacteria (Schreurs, 1992).

While many planktonic algae are grazed by copepods, daphnids and protozoa, cyanobacteria are not grazed to the same extent, and the impact of grazing by some specialised ciliates and rhizopod protozoans is usually not substantial.

Cyanobacteria are attacked by viruses, bacteria and actino-mycetes, but the importance of these natural enemies for the breakdown of populations is not well understood. Because they have few natural enemies, and their capacity for buoyancy regulation prevents sedimentation, the loss rates of cyanobacterial populations are generally low. Thus, their slow growth rates are compensated by the high prevalence of populations once they have become established.

Maximum growth rates are attained by most cyanobacteria at temperatures above 25 °C (Robarts and Zohary, 1987). These optimum temperatures are higher than for green algae and diatoms. This can explain why in temperate and boreal water bodies most cyanobacterial bloom during summer.

The ecological requirements of cyanobacteria discussed above vary between different species. As a consequence, different "ecostrategists" inhabit different types of water bodies. Understanding these ecostrategies may be useful for management, because it would help predicting which cyanobacteria can be expected to occur under certain conditions.

 

Predicting Cyanobacterial Occurrence and Blooms in Lakes and Reservoirs

Predicting and mitigating cyanobacterial blooms has not been as successful as predicting algal blooms. Much attention has been focused on managing cyanobacterial blooms. This effort has been more challenging and less successful than predicting algal blooms. If we could predict the onset of a cyanobacterial bloom we might be able to mitigate some of the negative effects and minimize the exposure to harmful toxic blooms. Predicting cyanobacterial blooms, however, is a complex and challenging question. Patterns from one lake often do not fit the patterns in a different lake. Thus far, there have been two main approaches to modeling cyanobacterial blooms in lakes:

  1. A process-based artificial neural network approach; and,
  2. Statistical approaches that are usually based on the chlorophyll:phosphorus relationship (Guven and Howard 2006). Artificial neural networks have been successful in predicting cyanobacterial blooms. For example, Maier et al. (1999, 2001) successfully forecasted Anabaena blooms in the River Murray, Morgan, Australia. Artificial neural networks are, however, extremely complex computational models that are constructed to mimic biological neural networks (Crawley 2007). Maier et al. (2001) pointed out that their model required a costly and intensive sampling effort that generated copious amounts of data (sampling was twice weekly for 7 years).

Statistical models such as linear regressions have also been used successfully to predict cyanobacterial blooms (e.g. Dillion and Rigler 1974, Onderka 2007). These models can be built based on existing data and are often more cost effective than artificial neural networks.

Early lake studies established the correlation between phosphorus and chlorophyll in lakes around the world (Sakamoto 1966, Dillon and Rigler 1974, Wetzel 2001). The positive relationship between chlorophyll (phytoplankton biomass) and total phosphorus laid the foundation of our current statistical models of phytoplankton blooms and helped shape our understanding of cyanobacterial blooms (Wetzel 2001, Havens 2008).

Recent studies redefine the chlorophyll-phosphorus relationship and incorporate additional variables into their models, including electrical conductivity, inorganic N (nitrite, nitrate), water temperature (Stanley et al. 2003) and total nitrogen/total phosphorus (TN:TP) ratio (Smith 1982). In a study using 228 lakes from the northern latitudes, Smith (1982) built a multiple regression model incorporating Sakamoto's equations on chlorophyll (Sakamoto 1966) and incorporated the TN:TP ratio. The regression had improved accuracy in predicting algal biomass, including cyanobacterial blooms (Smith 1982). However, Smith noted that moderate latitude, high nutrient lakes are not suitable for these models, based on variations in total nitrogen and total phosphorus.

Another limitation of the chlorophyll-phosphorus model is that it cannot predict phytoplankton blooms in phosphorus-rich lakes. In phosphorus-rich systems, nitrogen limitation is a better predictor of chlorophyll and has successfully predicted both algal and cyanobacterial blooms (Stanley et al. 2003). For instance, a study assessing water quality in Lake Manatee, Florida, found nitrogen limitation to be an important predictive variable in determining algal blooms (Stanley et al. 2003). Most phytoplankton biomass models built on the chlorophyll-phosphorus relationship have had success predicting algal blooms, but limited success in predicting cyanobacterial blooms (Downing et al. 2001, Guven and Howard 2006).

Luxury consumption of phosphorus and the ability to fix inorganic nitrogen contribute to the inaccuracy (Reynolds 1998, Ritchie et al. 2001). It is likely that other variables are contributing to the frequency and magnitude of cyanobacterial blooms, including light, temperature, turbidity (Reynolds 1985) and cyanobacteria buoyancy (Reynolds et al. 2002, Havens 2008).

A statistical model in Slovakia used three variables: total nitrogen, total phosphorus and temperature to predict cyanobacterial blooms. The model was successful in predicting cyanobacterial blooms in Liptovska Mara reservoir (Onderka 2007). The author also points out that the main goal of the study was to build a predictive model that would assist environmental managers and health officers in deciding when sampling of plankton should occur in order to save time and costly molecular analysis (Onderka 2007). Arguably, statistical models and artificial neural networks are narrowly constrained to the data being modeled, which limits their effectiveness when applied to other lake systems (Maier et al. 1997, Howard 2001, Guven and Howard 2006).

These models have had varying success and convey the complexity of predicting cyanobacterial blooms. It appears that a combination of factors are responsible for the occurrence of cyanobacterial blooms (Dokulil and Teubner 2000, Smith 2009). Currently, there is a need for the development of novel approaches to predicting cyanobacterial blooms that are cost effective, accessible to government agencies and water supply companies and timely, in mitigating the dangers of toxic blooms.

One such innovative approach was indirectly suggested by Smith (1983). Smith (1983) analyzed growing season data for 17 lakes throughout the world suggesting that a relative proportion of cyanobacteria in the epilimnetic phytoplankton is dependent on the epilimnetic ratio of total nitrogen to total phosphorous. Cyanobacteria tended to be rare when the ratio exceeded 29 to 1 by weight, suggesting that modifications to this ratio by controlled nutrient additions or removal may provide means by which lake water quality can be managed and blooms of cyanobacteria controlled.

Kostić et.al. (2016) in their study of Lake Vrutci in Serbia, conducted after a major bloom of Planktothrix rubescens in December of 2013, reported on the importance of the TN:TP ratio in controlling the bloom of Planktothrix rubescens during 2014 and 2015.

The potential of the TN:TP ratio to be used as the main predictor of the risk of cyanobacterial occurrence and blooms was further studied by Marjanović et al in 2016 as a part of the study of 31 lakes and reservoirs in Serbia using data set for these lakes for a period of 12 years from 2004 to 2015. The authors conclude that TN:TP ratio can be used as a reliable predictor of the risk of cyanobacterial blooms in lakes and reservoirs in Serbia.

 

TN:TP Ratio as a Predictor of Cyanobacterial Occurrence and Blooms in Serbia

The idea of TN:TP ratio being used as the main predictor of the risk of occurrence and blooms of cyanobacteria in lakes and reservoirs in Serbia, comes from the suggestion made by Smith (1983), that low TN:TP ratios promote dominance of cyanobacteria made on the basis of the analysis of data for 17 lakes worldwide (growing season data set for a single year for each of the studied lakes). Smith's analyses are summarized in Figure 1 showing the relationship between % cyanobacteria (biomass) and Mean epilimnetic TN:TP ratio.

 

fig01
Figure 1: Relationship between % cyanobacteria (biomass) and mean epilimnetic TN:TP ratio (after Smith, 1983)

 

We have analyzed data on abundance and biomass of algae (including % biomass of cyanobacteria, and concentration of total nitrogen and total phosphorous for 31 lakes for the period from 2004 to 2015. The number of samples taken from each lake during this period varied between the lakes and ranged from 56 to 252 samples. Samples and their analyses were carried out as part of the annual monitoring program of water quality conducted by the responsible state organization (Hydrometeorological Service of Serbia or Serbian Environmental Agency) and the data is regularly reported and is in the public domain. Samples are typically collected at a number of depths covering epilimnion, termocline and hipolimnion, and as a rule, close to the deepest point in the lake/reservoir. Of the 31 lakes/reservoirs studied, four had incomplete data sets and were excluded from further analyses.

 

fig02
Figure 2: Total number of samples analyzed for each lake/reservoir in the period from 2004 to 2015.

 

The available data was used to calculate average annual water column values for parameters of interest for each of the 27 lakes/reservoirs. (% Cyanobacteria and TN:TP ratio).

The results of our analysis are shown in Figure 3.

 

fig03
Figure 3: Relationship between % Cyanobacteria (biomass) and mean annual water column TN:TP ratio for 27 lakes/reservoirs in Serbia (Source of raw data: Annual Water quality monitoring reports of the State Hydrometeorological Service and Environmental Agency of Serbia, 2004 to 2015)

 

Close examination of the results presented in Figure 3 suggest the following:

Dominance and blooms of cynobacteria in lakes and reservoirs in Serbia do not occur if the mean annual water column TN:TP ratio is less than 29. This is in complete agreement with the findings of Smith (1983). It is extremely interesting that such an agreement seems to hold even for multi seasonal analyses and whole water column concentrations and not only epilimnetic concentrations, suggesting that "memory" and "history" of cyanobacterial occurrence play a significant role in the development of cyanobacterial blooms in lakes and reservoirs in Serbia.

Blooms of cyanobacteria in lakes/reservoirs in Serbia tend to occur only if cyanobacteria is present in the lake/reservoir in the previous growing season in excess of 20 % by weight.

The data suggests that the development of a cyanobacterial bloom is a multi-seasonal phenomenon in most lakes/reservoirs in Serbia. This seems to be a reasonable assumption since, in order for a cyanobacterial bloom to occur, it presumably requires the presence of a "SEED" population in the lake/reservoir.

Risk assessment of cyanobacterial blooms and interference with the water supply systems can therefore be associated with the whole water column TN:TP ratio and presence of cyanobacteria in the previous period (growing season or quarter) in general. On the basis of our findings, the risk assessment framework appropriate for the lakes and reservoirs in Serbia is presented in Figure 4 below.

 

fig04
Figure 4: Risk assessment framework appropriate for the assessment of the risk of cyanobacterial blooms in the lakes and reservoirs in Serbia.

 

Conclusions

Our findings along with the presented risk assessment framework are not sufficient to explain the cause and effect chain of events regarding the occurrence or absence of cyanobacteria in all lakes and reservoirs in Serbia since it is evident that many lakes and reservoirs with a TN:TP ratio < 29 are dominated by non cyanobacterial algae. This suggests that the structure of other trophic levels can alter phytoplankton response to nutrients (same as per findings of Smith, 1983).

Our data, nevertheless, suggests that lakes having whole water column TN:TP ratios > 29 will typically show very low proportions of cyanobacteria (Figure 3).

Practical implications of the above findings are that modification of TN:TP ratios can be a viable strategy for controlling the risk of cyanobacterial blooms. Methods and means for controlling TN:TP ratios need further study and elaboration.

 

References

Anagnostidis, K. and Komárek, J. (1985). Modem approach to the classification system of cyanophytes. 1 Introduction. Arch. Hydrobiol. Suppl. 71, Algological Studies, 38/39, 291- 302.

Castenholz, R.W. (1973). Ecology of blue-green algae in hot springs. In: N.G. Carr and B.A. Whitton [Eds] The Biology of Blue-Green Algae. Blackwell Scientific Publications, Oxford, 379-414.

Cohen-Bazire, G. and Bryant, D.A. (1982). Phycobilisomes: composition and structure. In: N.G. Carr and B.A. Whitton [Eds] The Biology of Cyanobacteria. Blackwell Scientific Publications, Oxford.

Crawley, M.J. (2007). The R Book. John Wiley and Sons, Ltd, England.

Dillon, P.J., and Rigler, F.H. (1974). The phosphorus-chlorophyll relationship in lakes. Limnology and Oceanography, 19 (5): 767-773.

Dokulil, M.T., and Teubner, K. (2000). Cyanobacterial dominance in lakes. Hydrobiologia, 438: 1-12.

Downing, J.A., Watson, S.B., and McCauley, E. (2001). Predicting cyanobacteria dominance in lakes. Canada Journal of Aquatic Science, 58: 1905-1908

Fay, P. 1965 Heterotrophy and nitrogen fixation in Chlorogloea fritschii. J. Gen. Microbiol. 39, 11-20.

Geitler, L. (1932). Cyanophyceae. In: L. Rabenhorst [Ed.] Kryptogamen-Flora. 14. Band. Akademische Verlagsgesellschaft, Leipzig, 1196 pp.

Gorham, P.R. and Carmichael, W.W. (1988). Hazards of freshwater blue-green algae (Cyanobacteria). In: C.A. Lembi and J.R. Waaland. [Eds] Algae and Human Affairs.

Guven, B., and Howard, A. (2006). A review and classification of the existing models of cyanobacteria. Progress in Physical Geography, 30 (1): 1-24.

Havens, K. E. (2008). Cyanobacteria blooms: effects on aquatic ecosystems. In H. H. Kenneth, editor, Cyanobacterial Harmful Algal Blooms: State of the Science and Research Needs. Springer-Verlag, New York, USA.

Howard, A. (2001). Modeling movement patterns of the cyanobacterium, Microcystis. Ecological Applications, 11(1): 304-310.

Ingrid Chorus and Jamie Bartram (1999). Toxic Cyanobacteria in Water: A guide to their public health consequences, monitoring and management, E & FN Spon, An imprint of Routledge, London and New York, 11 New Fetter Lane, London EC4P 4EE, © 1999 WHO

Jüttner, F. (1987). Volatile organic substances. In: P. Fay and C. Van Baalen [Eds] The Cyanobacteria. Elsevier, Amsterdam, 453-469.

Källqvist, T. (1981). Hydroecological field experiment 1981. Incubation of Natural Phytoplankton in Lake Gjersjøen. Norwegian Institute for Water Research, F-80402, Oslo, 21 pp

Kol, E. (1968). Kryobiologie. I. Kryovegetation. In: H.J. Elster and W. Ohle [Eds] Die Binnengewässer, Band XXIV. E. Schweizerbart'sche Verlagsbuchhandlung, Stuttgart, 216 pp.

Kostić et.al. (2016). Drivers of Phytoplankton Blooms in the Vrutci Reservoir During 2014-2015 and Implications for Water Supply and Management, Water Research and Management, 6(3), 3-12

Laamanen, M. (1996). Cyanoprokaryotes in the Baltic Sea ice and winter plankton. Arch. Hydrobiol. Suppl. 117, Algological Studies, 83, 423-433.

Linné, C. 1753 Species Plantarum. Tom II, Stockholm, 561-1200. Mann, N.H. and Carr, N.G. [Eds] (1992). Photosynthetic Prokaryotes. Biotechnology Handbooks, Volume 6, Plenum Press, London, 275 pp.

Maier, H.R., Dandy, G.C., and Burch, M.C. (1998). Use of artificial neural networks for modeling cyanobacteria Anabaena spp. In the River Murray, South Australia. Ecological Modeling 105: 257-272.

Maier, H.R., Sayed, T., and Lence, B.J. (2001). Forecasting cyanobacterium Anabaena spp. In the River Murray, South Australia, using B-spline neurofuzzy models. Ecological Modeling 146: 85-96.

Marjanović P. et.al. (2016). The Study of the Protection of Reservoirs in Serbia: Analysis of the Potential for the Occurance and Bloom of Cyanobacteria in Reservoirs in Serbia and Possibilities for Forecasts, Ministry of Agriculture and Environment of Serbia and Jaroslav Černi Institute, Belgrade 2016

Onderka, M. (2007). Correlations between several environmental factors affecting the bloom events of cyanobacteria in Liptovska Mara reservoir (Slovakia)- a simple regression model. Ecological Modeling, 209 (2-4): 412-416.

Onderka, M. (2007). Correlations between several environmental factors affecting the bloom events of cyanobacteria in Liptovska Mara reservoir (Slovakia)- a simple regression model. Ecological Modeling, 209 (2-4): 412-416.

Reynolds, C.S. (1987). Cyanobacterial waterblooms. In: P. Callow [Ed.] Advances in Botanical Research, 13, Academic Press, London, 17-143. Reynolds, C.S. 1997 Vegetation Processes in the Pelagic. A Model for Ecosystem Theory. Ecology Institute, D-21385 Oldendorf/Luhe, ISSN 0932-2205.

Reynolds, C.S. (1998). What factors influence the species composition of phytoplankton in lakes of different trophic status? Hydrobiologia 369(370): 11-26.

Reynolds, C.S., Huszar, V., Kruk, C., Naselli-Flores, L., and Melo, S. (2002). Towards a functional classification of freshwater phytoplankton. Journal of Plankton Research, 24 (5): 417-428

Robarts, R.D. and Zohary, T. (1987). Temperature effects on photosynthetic capacity, respiration, and growth rates of bloom-forming cyanobacteria. N.Z. J. Mar. Freshwat. Res., 21, 391-399.

Sakamoto, M. (1966). Primary productivity by the phytoplankton community in some Japanese Lakes and its dependence on lake depth. Archiv fur, Hydrobiologie, 62: 1- 28.

Schreurs, H. (1992). Cyanobacterial dominance, relation to eutrophication and lake morphology. Thesis, University of Amsterdam.

Skulberg, O.M. 1994 Oscillatoialean cyanoprokaryotes and their application for algal culture technology. Arch. Hydrobiol. Suppl. 105, Algological Studies, 75, 265-278.

Smith, V. H. (1983). Low Nitrogen to Phosphorous Ratios Favour Dominance by Blue-Green Algae in Lake Phytoplankton, Science, September 1983, Vol 221, No. 4611, 669-671

Smith, V.H. (1982). The Nitrogen Phosphorus dependence of algal biomass in lakes: an empirical and theoretical analysis. Limnology and Oceanography, 27(6): 1101- 1112.

Smith, V.H., and Schindler, D.W. (2009). Eutrophication science, where do we go from here? Trends in Ecology and Evolution, 24(4): 201-207.

Staley, J.T., Bryant, M.P., Pfennig, N. and Holt, J.G. [Eds] (1989). Bergey's Manual of Systematic Bacteriology. Vol. 3, Williams & Wilkins, Baltimore. Stewart, W.D.P. 1973 Nitrogen fixation by photosynthetic microorganisms. Ann. Rev. Microbiol., 27,283-316.

Stanley, C.D., Clarke, R.A., McNeal, B.L. and MacLeod, B.W. (2003). Relationship of chlorophyll a concentration to seasonal water quality in Lake Manatee, Florida. The Soil and Water Science Department, Florida Cooperative Extension, IFAS, University of Florida, http://edis.ifas.ufl.edu. Accessed February 2010.

Van Landingham, S.L. (1982). Guide to the Identification, Environmental Requirements and Pollution Tolerance of Freshwater Blue-Green Algae (Cyanophyta). United States Environmental Protection Agency, Cincinnati, Ohio, 341 pp.

Van Liere, L. and Mur, L.R. (1979). Chapter 9. Some experiments on the competition between a green alga and a cyanobacterium. In: L. Van Liere, Thesis, University of Amsterdam.

Van Liere, L., Mur, L.R., Gibson, C.E. and Herdman, M. (1979). Growth and physiology of Oscillatoria agardhii and some related species, a survey. Dev. Hydrobiol., 2, 67-77.

Vaucher, J.P. (1982). Historie des Conferves déau douce. Geneva. OECD Eutrophication of Waters, Monitoring, Assessment and Control. Organisation for Economic Co-operation and Development, Paris.

Walsby, A.E. (1981). Cyanobacteria: planktonic gas-vaculated forms. In: M. Star, H. Stolp, A. Balowes and H.G. Schlegel [Eds] The Prokaryotes. Springer Verlag, New York, 224- 235.

Walsby, A.E. (1987). Mechanisms of buoyancy regulation by planktonic cyanobacteria with gas vesicles. In: P. Fay and C. Van Baalen [Eds] The Cyanobacteria. Elsevier, Amsterdam, 377-414.

Wetzel, R.G., (2001). Limnology, Lakes and River Ecosystems, Third Edition. Elsevier, Academic Press, 00-110178.

Whitton, B.A. (1973). Freshwater plankton, In: G.E. Fogg, W.D.P. Stewart, P. Fay and A.E. Walsby [Eds] The Blue-Green Algae. Academic Press, London, 353-367.