Medium Sized Efficacy and Manual Delivery Proof of Concept

Medium Sized Efficacy and Manual Delivery Proof of Concept

15:46 30 May in Field Trail

Place: Greensboro AL, USA

Date: Aug.-Sep. 2015


This trial was intended to investigate whether a single treatment with Lake Guard White™ can trigger a sufficient and long-lasting collapse of the cyanobacterial population in large water bodies and to confirm the usability of the Lake Guard™ when applied manually.

Trial Setup

A field trial was conducted at Greensboro, AL (32°37’49.30″N/87°41’26.30″W) beginning on August 29, 2015. The trial was performed in five rainwater-filled earthen levee ponds (originally dug for catfish cultivation, 1.5-2.0 m deep), that had been out of use since 2007. Ponds were selected randomly. Treatments were applied to three ponds: pond H4 (10,100 m2); pond H5 (10,200 m2); and pond H7 (12,200 m2), while pond H1 (9,600 m2) and SP (2,800 m2), served as controls. Treatment consisted of a single application of 5.6 g/m2 of Lake Guard White™ (=5.0 g/m2 of the active compound NaDCC, Braun and Harel, 2013): 56 kg of Lake Guard White™ was applied onto ponds H4 and H5 whereas 70 kg was applied onto pond H7. The compound was manually dispersed from the northern side of the treated ponds and was driven across the pond within 30 min of dispersal by a moderate northerly wind. Each pond was sampled by a submerged sonde (at 10cm depth) at 5 fixed locations in each pond during the morning hours (07:00-11:00). All samples were obtained in a consistent order (SP, H1, H4, H5, and H7). Surface water from all ponds was sampled 1 hour prior to treatment (set as time ‘0’), 2 hours post treatment, and then daily during the following 3 days. Three additional samples were obtained on days 5, 10 and 16 from treatment.

Initial analysis revealed the presence of cyanobacterial cells identified as belonging to the genera Anabaena, Dactylococcopsis, Coelosphaerium, Cylindrospermopsis, Pseudoanabaena, Merismopedia, and Oscillatoria. Evidently, each pond was entirely different from the others in terms of initial phytoplankton and cyanobacterial composition, vegetation, organic load and even geological composition. Microscopic cell-count revealed initial cell densities of ~5X105 cyanobacteria/ml in ponds SP, H1 and H5, while cyanobacterial densities in ponds H4 and H7 were an order of magnitude lower.

Trial Protocol

A YSI-Exo-1 probe (Xylem Ltd, Yellow Springs, Ohio, USA) was used to measure temperature, pH, dissolved oxygen (DO), specific conductivity (SpC), Chlorophyll-a (“Chl-a”, as an indicator for total phytoplankton) and Phycocyanin (“PC”, as a proxy for total cyanobacteria). Chl-a and PC were calibrated according to manufacturer’s instructions using a wild strain of Microcystis sp.

Total chlorine was determined by ‘Total Chlorine Pocket Tester’ (LaMotte 1742, LaMotte MD, US.).

Total particulate matter was assessed using a Clogging Potential Meter (Israel Water Works Association, Israel) with a 33 µm sieve filter. Clogging potential was measured for a total of 6 days, beginning 1 day prior to treatment application. This device measures the time it takes to create 5 m differential pressure across a 33 µm sieve filter at a constant flow rate of 10 liters per minute (Feldlite and Yechiely, 2011; Milstein and Feldlite, 2014). In principle, the longer it takes the filter to become clogged – the better the water quality is.

Data analysis

Relative values were used for data analysis in order to overcome the heterogeneity of the studied ponds and for standardization of the experimental data. Relative values were calculated and represent the individual changes occurring in each of the ponds after treatment.

Trial Results

Algaecide dispersal and chlorine measurements

Following manual dispersal, algaecide particles were driven across the pond by a mild northerly wind (Fig. 1A) forming a distribution pattern closely matching that of the floating cyanobacterial cells (Fig. 1B). Chlorine levels, measured in the water containing the floating algaecide particles one hour following application, were found to be below detection levels of 0.01 ppm, indicating that all available chlorine had interacted with the organic material in the ponds’ water. The empty coating of the algaecide formulation had disintegrated in the water within 24-48 hours.


Fig. 1. Photographs of the spatial formation of Lake Guard White™ particles in pond H4 (A). View to South-East, arrow indicates the change in water surface tension due to the hydrophobic nature of the algaecide particles. (B) Downstream accumulation of Lake Guard White™ particles two hours after treatment and their formation along the floating clusters of cyanobacteria.

Chlorophyll-a and phycocyanin measurements

A substantial reduction in both Chl-a and PC concentrations was measured in all treated ponds within 48 hours from algaecide application (Fig. 2A). PC concentrations, indicating total cyanobacterial abundance, were reduced by an average of 56.6% and 60.3% relative to time ‘0’ in samples obtained at 2 and 24 hours following treatment, respectively (Fig. 2A-B). Further decreases in PC levels averaging 83.9% and 79.01% were measured 48 and 72 hours after treatment, respectively. This trend persisted throughout the experiment (16 days after treatment) in two of the treated ponds (Fig. 2B). A sudden rise in PC concentrations to pre-treatment levels was observed in pond H4 on day 10, with a subsequent reduction observed on day 16 (Figs. 2A-B). Following initial reduction in Chl-a content in all treated ponds (Figs. 2A and 2C), a significant increase was noted starting at 72 hours after treatment, with peak concentrations measured on day 10 (Figs. 2A and 2C).


Fig. 2. (A) Changes in average PC and Chl-a readings between treated and untreated ponds. Data is normalized to day 0 (two hours before application) according to equations 1 and 2. Inlet: a zoom-in on the first two days of the trial to emphasize %PC decline in treated ponds.

(B) Post-treatment PC content reduction calculated based on averaged data according to equation 1.

(C) Post-treatment Chl-a content reduction calculated based on averaged data according to equation 2.

Bars represent standard error.


Other water measurements

In distinction to the substantial reduction in average PC levels measured in the treated ponds, average pH levels in treated ponds (Fig. 3A) remained constant, ranging between pH 8.36 to pH 8.0 during the first 5 days of the trial. PH levels of 7.7 and 8.2 were measured on days 10 and 16, respectively. Average pH in the untreated ponds was higher measuring pH 9.3 on days 1 and 2, falling to 8.7 on days 3 and 5 followed by measuring of pH 8.5 and pH 9.0 on days 10 and 16, respectively. Average DO concentrations measured in the water were relatively similar for both treatments with a relative low standard error between the ponds (Fig. 3B). The same trend was also recorded for average SpC values at ~140-175 Siemens/cm during the first 10 days of the trial (Fig. 3C). Pond water temperature (Fig. 3D) rose from 27.1°C to 31.8°C during the first 5 days of the trial, followed by water temperatures of 28.9°C and 26.0°C recorded on days 10 and 16, respectively.

The single treatment with the Lake Guard White™ substantially improved the average pond water filterability for at least 5 days. Compared to measurements obtained 1 day prior to treatment, average time to clogging (note Fig. 4A) in the treated ponds increased by 50%-100% following treatment (Fig. 4B). Measurements were discontinued on day 5 due to technical constraints.

Fig. 3. Changes in average values of (A) pH;

(B) dissolved oxygen (DO);

(C) Specific conductivity (SpC) – in treated and untreated ponds throughout the trial. Data (A-C) was normalized to day 0. Bars represent standard error.

(D) Raw data for water temperature throughout the trial.

Fig. 4. The difference in total particulate matter in the water between treated and untreated ponds was determined by (A) a Clogging Potential Meter equipped with 33 µm sieve filter (A, inner photograph).

(B) Data from all ponds was averaged and normalized to day 0. By day 2, the average time it took the filter to become clogged in treated ponds was 50% longer than in the controls.



A single treatment with Lake Guard™ triggers collapse of the cyanobacterial population

The aim of this trial was to investigate whether a single treatment of Lake Guard White™ at a concentration of 5 g/m2 NaDCC can trigger a long-lasting reduction in cyanobacterial populations as was previously observed in small scale, 40 cm2 enclosures. The large surface area covered in this field trial (a total of 45,000 m2 of water surface) mandated the use of a measuring method that would be reliable and allow easy sampling in multiple sampling locations. To that end, a YSI-Exo-1 probe was used to measure multiple water indicators, including phycocyanin (PC), a pigment specific to cyanobacteria and a good indicator for cyanobacterial biomass (Loisa et al., 2015).

Our results confirm an immediate effect of the algaecide on cyanobacterial cell density in all treated ponds (Figs. 2A-B) with no such reduction observed in the untreated ponds. The intensity of the effect varied between treated ponds, likely due to the different starting conditions in each pond. For ponds H4 and H7, PC content was at or below the detection level at days 2 and 3 of the 7 sampling days (Fig. 2B). A transient and short lasting increase in PC content was observed on day 3 in pond H7 and on day 10 in pond H4. Such sharp fluctuations are characteristic of microbial communities with low diversity, where viral activity may check the rapid growth of specific bacterial populations through a kill-the-winner mechanism (Shapiro et al., 2009). This phenomenon should be further investigated in designated trials.

Treatment was evidently less effective in pond H5, where initial cyanobacterial concentrations were an order of magnitude higher than in the two other treated ponds. Nevertheless, a 50% decrease in PC content was measured immediately following treatment which lasted until the end of the sampling period (Figs. 2A-B). This result is in agreement with observations from preliminary experiments (see supplementary information), where cyanobacterial blooms with higher cell densities were shown to be less susceptible to the algaecide treatment. Based on these results, we propose to define a cell density of ~104 cyanobacteria/ml as an upper threshold for the floating algaecide intervention under a 5 g/m2 treatment regime. This definition also corresponds with the WHO threshold for safe recreational exposure (WHO, 2003). A treatment at such an early stage of cyanobacterial blooms is expected to be safer in terms of toxin release (by lysed cells), cost-effective, have a long-lasting effect and minimize the environmental impact of both bloom and treatment.

The observed change in particulate matter concentrations (Fig. 4B) further supports the strong and lasting effects of the algaecide treatment. As cyanobacteria are known to secrete high amounts of polysaccharides, possibly as a means to alter their environment (Jenkinson and Sun, 2010; Harel et al. 2012), it is expected that a removal of a large proportion of their biomass would significantly improve water filterability. This finding further supports the notion of a fundamental difference between the phytoplankton communities in treated and untreated ponds. Since this is the first observation of its kind, additional investigation is needed.

NaDCC was found to fully release its chlorine content within 60 minutes of application and to interact with the existing organic load in the water, after which total-chlorine could not be detected in the water (<0.01 ppm). In contradiction to in-vitro tests done in purified water (i.e. Clasen and Edmondson, 2006) this rapid interaction released all bound chlorine regardless of the pH range in the ponds (pH 8.0 to pH 8.34, Fig. 3A). Cumulative experience with hypochlorite-releasing formulas has brought us to the understanding that the effective concentration of the free-chlorine in the water is determined by the following factors: (i) the arithmetic ratio of the actual NaDCC quantity and the overall volume of water in each pond which translated here into 1.5 to 2 ppm of total chlorine; (ii) the time it takes each granule of algaecide to release its chlorine content to the water; (iii) the organic load available in the water. Factoring all of the above leads to the conclusion that the treatment had not achieved a ‘total bleaching’ effect (complete eradication of living organisms). In fact, the initial theoretical value of total chlorine in the entire pond is comparable with free chlorine values permitted in drinking water. The rapid drop in total-chlorine concentrations to below detectable levels (of 0.01ppm) is related to the formulation structure, which causes hypochlorite ions to immediately interact with the organic load present at the water surface once they are released from the algaecide granules. Therefore, the decline observed in cyanobacterial levels in the pond trial as well as at the 40 cm2 enclosure trials (LINK TO FIGURES ABOVE: 1-3) is disproportional to the actual amount of chlorine used and must not be associated only with a chemical effect but also with a biological process.


Lake Guard™ efficacy enhanced by the secretion of a diffusible cyanobacterial signaling molecule

The significant decline in cyanobacterial populations in treated ponds cannot be linked to biotic or abiotic parameters such as DO, SpC or weather conditions (Figs. 3A, 3C and 3D), as all parameters in both treated and untreated ponds followed the same trends. These results are in agreement with a previous 40 cm2 enclosure trial and present additional evidence for the negligible environmental effect of the chlorine. Changes in pH levels did not vary significantly throughout the trial (Fig. 3A) which may be related to the geological character of the local soil, also known as “black prairie soil”, known for its strong buffering trait. Consequently, the water in the ponds was probably well buffered, masking the expected decrease in pH levels at treated ponds as was previously observed in the 40 cm2 enclosure trials.

Considering chlorine levels were not sufficient to bleach the entire water column, the extensive cyanobacterial cell death (Figs. 2) may therefore be linked to the oxidative stress at the water surface induced by the slow-releasing Lake Guard™ formula. Results from preliminary trials performed at Mishmar Ayalon, Israel ( suggested that the directly affected cyanobacteria secreted specific diffusible signaling molecules. These cues may have propagated throughout the water body and initiated an auto-catalytic response triggering cell-death in naïve cyanobacterial cells throughout the water column. Such a mechanism could account for the subsequent domination of non-harmful eukaryotic algae (as further elaborated hereafter) which were not affected by the auto-catalytic cell-death signals secreted by neighboring cyanobacteria.

Programmed cell death in cyanobacteria has been extensively investigated (Berman-Frank et al., 2004; Ross et al., 2006; Bar-Zeev et al., 2013). The signal amplification concept for cyanobacterial cell-death is also in agreement with Sedmak et al. (2008) and Cohen et al. (2014) who showed that stressed cyanobacterial cultures secreted toxic compounds or “signals” that initiated a self-regulated death cascade within their own populations. Both studies proposed that this type of induction mechanism might be connected to naturally occurring, spontaneous mass cyanobacterial collapse events that can be observed in nature at the end of every bloom cycle (Fallon and Brock, 1979).

Indications for green algae succeeding cyanobacteria

The opposing trends of PC and Chl-a levels in the treated ponds (Figs. 2A ) suggest that while cyanobacteria levels were declining in the treated ponds, eukaryotic photosynthetic microorganisms (containing Chl-a) were occupying the resulting “vacuum”, possibly by benefiting from the elimination of competition as well as from the nutrients released by the lysis of cyanobacterial cells. This observation is in agreement with results of enclosure studies performed by us earlier (see Supplementary Material) and work done by Matthijs et al. (2012) suggesting increased sensitivity of cyanobacteria to oxidative stress.

Put together, these findings suggest that the proliferation of eukaryotic photosynthetic microorganisms following the cyanobacterial decline may prevent the return of cyanobacterial populations (Fig. 2A). Matthijs et al. (2012) treated a shallow lake with liquid hydrogen peroxide and achieved up to seven weeks of reduction in cyanobacteria levels with a corresponding rise in levels of green algae, through what appears to be a similar mode of succession.

Environmental implications of chlorine discharge into large water bodies

While the concept of releasing free-chlorine into large bodies of water may seem to be an unorthodox measure, it is, in fact, a common practice in power plant cooling-water systems. These facilities have been authorized by regulators to treat their cooling-water systems with chlorine (in order to avoid biofouling) which is later being discharged in massive quantities to adjunct water bodies. This method is regarded as ‘the best available method’ in terms of biocidal efficiency and cost effectiveness (Hergott et al., 1978). The US Code of Federal Regulations (40 CFR 423) dictates limitations on the concentration of free-chlorine and the permissible amount of chlorinated water discharge to be released into a water body per day per reactor: A power plant with a standard daily flow of five million cubic liters can release up to 200,000 kg of free-chlorine into the adjacent pond on an annual basis (see Table 3.1 in Pacey et al., 2011). Comprehensive environmental studies that have been conducted by various environmental protection agencies showed that this mass discharge of chlorine had minor adverse effects on fauna and flora (e.g. Brungs, 1973; Hergott et al., 1978; Sung et al., 1978; Pacey et al., 2011; Ma et al., 2011). Moreover, studies also showed that residual chlorine byproducts from these practices, such as trihalomethanes, were found below harmful levels (Hollod and Wilde, 1982; Jenner et al., 1997; Jenner and Wither, 2011).


The results presented here clearly show that a single treatment with 5.6 g/m2 Lake Guard White™ at early stages of a cyanobacterial bloom triggered a massive collapse of cyanobacterial populations in the treated water bodies. This collapse was followed by the succession of non-harmful phytoplankton in the ecological niche which, in turn, transiently prevented the cyanobacterial populations from rebounding. The amount of chlorine released by sporadic surface treatment of the algaecide during early bloom events is expected to be well within the limitations set by multiple environmental protection agencies for the discharge of free-chlorine into water bodies used for power-plant cooling systems. This conclusion is underscored when considering that existing regulatory approvals are based on economic grounds resulting in chlorine discharge into “healthy” water bodies, whereas our suggested algaecidal application is designed to cure “sick” water bodies at a significant risk of a fully developed toxic cyanobacterial bloom.



Bar-Zeev,E., Avishay,I., Bidle,K.D., and Berman-Frank,I. (2013) Programmed cell death in the marine cyanobacterium Trichodesmium mediates carbon and nitrogen export. ISME J 7 (12), 2340-2348.

Berman-Frank,I., Bidle,K.D., Haramaty,L., and Falkowski,P.G. (2004) The demise of the marine cyanobacterium, Trichodesmium spp., via an autocatalyzed cell death pathway. Limnology and Oceanography 49 (4), 997-1005.

Bowling,L., Ryan,D., Holliday,J., and Honeyman,G. (2013) Evaluation of in situ fluorometry to determine cyanobacterial abundance in the Murray and Lower Darling Rivers, Australia. River Research and Applications 29 (8), 1059-1071.

Braun, Sergei and Harel, Moshe. A method of controlling water surface inhabiting pests.  2013. WO2015001563A1, SHA’KED MICROBIAL SOLUTIONS LTD.  

Brungs,W.A. (1973) Effects of residual chlorine on aquatic life. Water Pollution Control Federation 45 (10), 2180-2193.

Clasen,T. and Edmondson,P. (2006) Sodium dichloroisocyanurate (NaDCC) tablets as an alternative to sodium hypochlorite for the routine treatment of drinking water at the household level. International Journal of Hygiene and Environmental Health 209 (2), 173-181.

Cohen,A., Sendersky,E., Carmeli,S., and Schwarz,R. (2014) Collapsing aged culture of the cyanobacterium Synechococcus elongatus produces compound(s) toxic to photosynthetic organisms. PLoS ONE 9 (6), e100747.

Fallon,R.D. and Brock,T.D. (1979) Lytic Organisms and Photooxidative Effects: Influence on Blue-Green Algae (Cyanobacteria) in Lake Mendota, Wisconsin. Appl.Environ.Microbiol. 38 (3), 499-505.

Fan, J. (2013). Evaluating the effectiveness of various control and water treatment processes on the membrane integrity and toxin fate of cyanobacteria. PhD Thesis from the Univerity of Adelaid, Australia. LINK

Feldlite, M. and Yechiely, Z. (2011) On-line monitoring of clogging potential in irrigation systems p. 89-90.

Garcia-Villada,L., Rico,M., Altamirano,M., Sanchez-Martin, Lopez-Rodas,V., and Costas,E. (2004) Occurrence of copper resistant mutants in the toxic cyanobacteria Microcystis aeruginosa: characterisation and future implications in the use of copper sulphate as algaecide. Water Research 38 (8), 2207-2213.

Harel,M., Weiss,G., Daniel,E., Wilenz,A., Hadas,O., Sukenik,A., Sedmak,B., Dittmann,E., Braun,S., and Kaplan,A. (2012) Casting a net: fibres produced by Microcystis sp. in field and laboratory populations. Environmental Microbiology Reports 4 (3), 342-349.

Hergott, Steven J., Jenkins, David, and Thomas, Jerome F. Power Plant Cooling Water Chlorination in Northern California. Water Pollution Control Federation 50[11], 2590-2601. 1978.

Hollod,G. and Wilde,E. (1982) Trihalomethanes in chlorinated cooling waters of nuclear reactors. Bull.Environ.Contam.Toxicol. 28 (4), 404-408.

Jenkinson, I. R. and Sun, J. Thalassorheology, algal blooms and vertical carbon flux. Pagou, P. and Hallegraeff, G. Proceedings of the 14th International Conference on Harmful Algae.International Society for the Study of Harmful Algae and Intergovernmental Oceanographic Commission of UNESCO 2013 . 2010.

Jenner, H. A and Wither, A. W. Chlorination by products in power station cooling water. 009. 2011.  British Energy Estuarine & Marine Studies.

Jenner,H.A., Taylor,C.J.L., van Donk,M., and Khalanski,M. (1997) Chlorination by-products in chlorinated cooling water of some European coastal power stations. Marine Environmental Research 43 (4), 279-293.

Loisa,O., Kaaria,J., Laaksonlaita,J., Niemi,J., Sarvala,J., and Saario,J. (2015) From phycocyanin fluorescence to absolute cyanobacteria biomass: an application using in-situ fluorometer probes in the monitoring of potentially harmful cyanobacteria blooms. Water Practice and Technology 10 (4), 695-698.

Lurling,M., Waajen,G., and de Senerpont Domis,L. (2015) Evaluation of several end-of-pipe measures proposed to control cyanobacteria. Aquat Ecol 1-21.

Ma,Z., Lin,H., Gu,X., and Xu,Z. (2011) Effects of residual chlorine on the mortality, grazing and respiration of Labidocera euchaeta (Copepoda). Acta Oceanol.Sin. 30 (2), 96-102.

Matthijs,H.C.P., Visser,P.M., Reeze,B., Meeuse,J., Slot,P.C., Wijn,G., Talens,R., and Huisman,J. (2012) Selective suppression of harmful cyanobacteria in an entire lake with hydrogen peroxide. Water Research 46 (5), 1460-1472.

Milstein,A. and Feldlite,M. (2014) Relationships between clogging in irrigation systems and plankton community structure and distribution in wastewater reservoirs. Agricultural Water Management 140 79-86.

Pacey, N, Beadle, I., Heaton, A., and Newsome, L. Chemical discharges from nuclear power stations: historical releases and implications for Best Available Techniques (BAT).  2011. Bristol, Environment Agency, Horizon House, Deanery Road.

Paerl,H.W. and Otten,T.G. (2013) Blooms Bite the Hand That Feeds Them. Science 342 (6157), 433-434.

Papadimitriou, T., Kagalou, I., Stalikas, C., Pilidis, G., & Leonardos, I. D. (2012). Assessment of microcystin distribution and biomagnification in tissues of aquatic food web compartments from a shallow lake and evaluation of potential risks to public health. Ecotoxicology, 21(4), 1155-1166.

Roegner,A.F., Brena,B., Gonzalez-Sapienza,G., and Puschner,B. (2014) Microcystins in potable surface waters: toxic effects and removal strategies. Journal of Applied Toxicology 34 (5), 441-457.

Ross,C., Santiago-Vazquez,L., and Paul,V. (2006) Toxin release in response to oxidative stress and programmed cell death in the cyanobacterium Microcystis aeruginosa. Aquatic Toxicology 78 (1), 66-73.

Sedmak,B., Carmeli,S., and Elersek,T. (2008) “Non-Toxic” Cyclic Peptides Induce Lysis of Cyanobacteria an Effective Cell Population Density Control Mechanism in Cyanobacterial Blooms. Microbial Ecology.

Shapiro, O. H., Kushmaro, A., & Brenner, A. (2010). Bacteriophage predation regulates microbial abundance and diversity in a full-scale bioreactor treating industrial wastewater. The ISME journal, 4(3), 327-336.

Sung, R., Strehler, D., and Thorne, C. Assessment of the effects of chlorinated seawater from power plants on aquatic organisms. Final task report, May-September 1978.  1978.  TRW Environmental Engineering Div., Redondo Beach, CA (USA).

Takeda,T., Yokota,A., and Shigeoka,S. (1995) Resistance of photosynthesis to hydrogen peroxide in algae. Plant and Cell Physiology 36 (6), 1089-1095.

Tamoi,M., Murakami,A., Takeda,T., and Shigeoka,S. (1998) Acquisition of a new type of fructose-1,6-bisphosphatase with resistance to hydrogen peroxide in cyanobacteria: molecular characterization of the enzyme from Synechocystis PCC 68031. Biochimica et Biophysica Acta (BBA) – Protein Structure and Molecular Enzymology 1383 (2), 232-244.

Torbick,N. and Corbiere,M. (2015) A multiscale mapping assessment of Lake Champlain cyanobacterial harmful algal blooms. International Journal of Environmental Research and Public Health 12 (9), 11560-11578.

WHO. (2003). Guidelines for safe recreational water environments: Coastal and fresh waters. Vol. 1. World Health Organization.

Wolf, David, and H. Allen Klaiber. Bloom and Bust: Toxic Algae’s Impact on Nearby Property Values. No. 235159. Agricultural and Applied Economics Association, 2016.

Wynne,T.T. and Stumpf,R.P. (2015) Spatial and temporal patterns in the seasonal distribution of toxic cyanobacteria in Western Lake Erie from 20022014. Toxins 7 (5), 1649-1663.