Because of the many characteristics of cyanobacteria, remote sensing provides an effective monitoring tool. Using satellite data, we can provide forecasts and early warning of the location and extent of algal blooms.
Figure 1. Warning sign for cyanobacteria. Image Credit: J. Graham, USGS.
NOAA, National Centers for Coastal Ocean Science
A swim in the summer in a lake is refreshing, unless the lake is green (Figure 1).åÊ Blooms of cyanobacteria, also called blue-green algae, are a risk to swimmers in at least 36 states and 50 countries [1].åÊ In 2002, a coroner determined that a boy in Wisconsin died from cyanobacteria toxins after going for a swim [2].åÊ åÊIn 2009-2010, cyanobacteria caused nearly half of the illness outbreaks in lakes reported to the U.S. Centers for Disease Control [3].åÊ And in 2012, dogs died from cyanobacterial toxins after playing in lakes in California, Maryland, Indiana, New York, and Oklahoma [4][5].åÊ The bloom can be costly, as well, if a community is not prepared; in 2013, a small town’s water plant on Lake Erie shut down for two days.åÊ These were all caused by blooms of cyanobacteria, one of a number of organisms that produce toxins leading to harmful algal blooms (HABs).
HABs are a general term for algae, or cyanobacteria, that contain toxic substances. Most HAB-forming species live in the ocean and are one-celled organisms called dinoflagellates.åÊ Cyanobacteria, which are photosynthetic bacteria, are the most common HAB organism in fresh or brackish water.åÊ They are ubiquitous—found in oceans lakes, ponds, geothermal pools and other places, as well as one of the oldest groups of organisms on Earth, with fossils in Australia that are 3.5 billion years old.
Figure 2. MODIS true color image of Lake Erie, Oct. 9, 2011. Bright green areas are scum formed by cyanobacteria near the end of the worst bloom in decades. Image Credit: R. Stumpf, NOAA.
While cyanobacteria may form mat-like layers (which led to the Australian fossils), they are most often planktonic, floating in the water, and usually present in low and harmless concentrations. Many planktonic cyanobacteria have vacuoles, which are storage areas in the cell.åÊ These vacuoles fill with carbohydrates as they photosynthesize during the day, and fill with gases overnight [6]. As a result, the cells can move up and down in the water, providing access to nutrients in at night, and light in the day.åÊ For several species, like those of Microcystis, the flotation leads to green scum seen on calm days, at a scale that may be seen via satellite (Figure 2).
Several common cyanobacteria can be toxic. Microcystis aeruginosa, produces toxins called microcystins; this species is one of the most common causes of cyanobacteria blooms in the United States, and it produces scum layers.åÊ Microcystin and other toxins, such as anatoxins and cylindrospermosin, are produced by species of several common genera including Anabaena,åÊ Planktothrix, Aphanizomenon, and Cylindrospermopsis [2]
Figure 3. Cyanobacteria (bright green) floating up in a sample, while diatoms (olive green) sink.
The toxicity should not be underestimated.åÊ Microcystin is a liver toxin, and the consequence of a toxic dose for people or animals is liver and kidney failure.åÊ In 1996, more than 50 patients at a dialysis center in Brazil died as a result of microcystin getting into the water supply [7].åÊ The World Health Organization (WHO) has recommended that drinking water should have no more than 1 part per billion (ppb) of microcystin, and recreational water for swimming poses a risk at 10 ppb.åÊ Over the last several decades, hundreds of dogs have died after exposure to cyanobacterial toxins, and this considers only the documented cases [5].åÊ For swimmers, the greatest risk comes from accidentally swallowing water; however the toxins can cause dermatitis and people or pets that get into such water should be promptly washed.
Eliminating cyanobacteria is difficult.åÊ Reducing the blooms requires reducing nutrient loads, especially from phosphorus.åÊ Otherwise, cyanobacteria grow best in nutrient-rich calm, warm water—conditions åÊthat occur in summer especially after a wet spring—because they are so small and can float [6].åÊåÊåÊ Other (harmless) algae, like diatoms, can grow rapidly, but are heavy and sink when the water is calm (Figure 3).åÊ When the water has excessive nutrients, especially phosphorus, the cyanobacteria keep growing, and in the summer they block light from the diatoms, becoming as dense as the nutrient load will allow.åÊ In the 1960s and 1970s, many rivers and lakes had dense blooms because of the lack of treatment of sewage and industrial waste.åÊ With laws (like the U.S. Clean Water Act), these sources were removed.åÊ The major source of phosphorus and nitrogen today is from runoff from non-point sources, such as farms, lawns, roadsand parking lots.
Many lakes are sources of drinking water, besides being popular for swimming and boating.åÊ Blooms that are floating on the surface can blow to one end of a lake, which means that the other side might be safe for swimming.åÊ Public water suppliers can manage the blooms, provided they have advance warning that a bloom is forming or coming.åÊ This includes filtering cells before treatment starts, and using activated charcoal and other treatments that reduce the toxins.åÊ These changes are highly effective, but expensive; for a small city, the additional treatments may cost $3,000 a day [4].åÊåÊåÊ Detecting these blooms as soon as possible, and monitoring and forecasting their development will help water suppliers respond when they need to, and not when they don’t, and help vacationers, and fishing and boating communities.åÊ Understanding the location and timing of the blooms also can lead to strategies for nutrient reduction.
Figure 4. Reflectance spectra of an intense Microcystis bloom over visible and near-infrared (NIR) wavelengths, with location of MERIS red and NIR bands. Blue light (400-500 nm) and red light (600-700 nm) are absorbed, leading to the common bright green color.
Cyanobacterial blooms typically are a problem at relatively high concentrations.åÊ The WHO guidelines for recreational exposure have estimate that cell concentrations equivalent to 10 åµg (micrograms) chlorophyll L-1 (about 20,000 cells L-1) pose some risk of irritation, and 50 åµg chlorophyll L-1 have a moderate risk [8].åÊ Scum is the greatest risk for swimmers.åÊ For water suppliers, as most microcystin is in the cells, there is a similar concern for total amount.åÊ If the cells are removed intact, subsequent treatment is easier; if they are lysed (broken) before removal, the toxin in the cells ends out dissolved in the water and requires more treatment.
Remote Sensing as a Monitoring Tool
Because of the many characteristics of cyanobacteria, remote sensing provides an effective monitoring tool.åÊ The blooms are bright because they scatter light strongly (because of the cell structure), they are often near the surface (during calm weather), and they have chlorophyll and phycocyanin.åÊ Phycocyanin absorbs orange-red light strongly with the greatest absorption at wavelengths about 620 nanometers (nm).åÊ Chlorophyll, while absorbing blue light, and some red light, absorbs especially strongly åÊnear 680 nm (in the red).åÊ Water absorbs near-infrared light quite strongly starting about 720 nm.åÊ The result is a distinctive spectra (Figure 4), with the absorption causing dips at around 620 nm, 680 nm, and strong reflectance around 700 nm. (In blue and green wavelengths below 500 nm, many pigments absorb light, including chlorophyll, carotenoids, and non-algal pigments like tannins and iron compounds).
Figure 5. Index of amount of cyanobacteria in Lake Erie from MERIS on Oct. 8, 2011, (the day before Figure 2). Medium corresponds to the moderate risk of WHO (of about 100,000 cells L-1). Image Credit: R. Stumpf , NOAA.
Satellites have provided an effective means of finding these blooms [9][13].åÊ One particularly effective sensor was the Medium Resolution Imaging Spectrometer (MERIS) that was on the European Space Agency’s (ESA) Envisat-1 satellite.åÊ MERIS had bands that overlay the absorption peaks in the red and near-infrared (Figure 4).åÊ The result is that the chlorophyll and phycocyanin patterns can be targeted.
One powerful method of finding these dips and peaks is to determine the curvature, which (in calculus) is the second derivative, a method used for several satellite algorithms [12].åÊ The amount of curvature around 680 nm gives the amount of chlorophyll, and the direction of curvature between 620 and 680 nm indicates the presence of phycocyanin [11].åÊ Using this information, the blooms can be identified and quantified (Figure 5).åÊ Other satellites, like the Moderate Resolution Imaging Spectro-radiometer (MODIS), have a 680 nm band, so they can indicate quantity of biomass.åÊåÊ The use of curvature algorithms has an additional advantage in that the blooms can be found without having to determine how much of the signal came from the atmosphere [12].åÊ Even on clear days, the atmosphere makes up 90 percent of all the light reaching the satellite; and the atmosphere does not produce dips or curves seen in visible water spectra.åÊ This means that looking at curvature in the spectrum can identify pigments, regardless of most atmospheric haze.åÊ Methods that avoid trying to calculate the atmospheric signal are more reliable for consistent monitoring of blooms.
Using the satellite data, we can provide forecasts and early warning of location and extent of the blooms.åÊåÊ The National Oceanic and Atmospheric Administration (NOAA) has conducted routine monitoring of Lake Erie for cyanobacterial blooms each summer for several years [13].åÊ MERIS åÊand MODIS have provided indices that show the severity of the bloom (Figure 6). Transport over several days can be forecast by moving the bloom with currents generated by a hydrodynamic model.åÊ Public water suppliers use the information to plan adjustments in water treatment.åÊ Health and parks departments can post (or remove) advisories from beaches as a result of the information.
![Figure 6. The maximum intensity of the bloom in Lake Erie determined by integrating the cyanobacteria index over 30 days [10]. A CI of 1 corresponds to about 1020 cells of Microcystis.](https://ieeeearthzine.wpenginepowered.com/wp-content/uploads/2014/03/figure-6.png)
Figure 6. The maximum intensity of the bloom in Lake Erie determined by integrating the cyanobacteria index over 30 days [10]. A CI of 1 corresponds to about 1020 cells of Microcystis.
![Figure 7. The relationship of spring discharge into western Lake Erie to severity of the bloom from [10]. Higher flow leads to larger phosphorus loads, which drive the blooms. This relationship has allowed annual predictions of bloom severity.](https://ieeeearthzine.wpenginepowered.com/wp-content/uploads/2014/03/figure-7.png)
Figure 7. The relationship of spring discharge into western Lake Erie to severity of the bloom from [10]. Higher flow leads to larger phosphorus loads, which drive the blooms. This relationship has allowed annual predictions of bloom severity.
With a longer data set, better understanding of nutrient loading and climate impacts can be made.åÊ While land management and control of runoff is the major factor in ultimately controlling nutrient loads, the variation between years depends on rainfall, which depends river discharge (Figure 7).åÊ For example, in 2011, the floods from a wet spring led to the worst bloom in decades for Lake Erie. åÊThe Maumee River basin is predominately agricultural, so strategies can be developed to reduce phosphorus during the spring, thereby reducing the severity of these toxic blooms [14].
Cyanobacteria produce toxic and nuisance blooms around the world.åÊ Routine monitoring to know where and when the blooms occur can reduce the expense of protecting people (and their pets).åÊ Satellite data will be a key part of that information.åÊ More observations will lead to a better understanding of the specific environmental factors that change the blooms; for example, the role of changing temperature in prolonging blooms, of winds in mixing the blooms through the water, and of the nutrient sources that drive the blooms.åÊ The monitoring can lead to plans that can reduce the frequency of the blooms and the risk of illness.
Acknowledgments
This work was partially funded by the NASA Applied Science Program announcement NNH08ZDA001N, Project # NNH09AL53I.
References
[1] J. L. Graham,åÊ K. A. Loftin, N. Kamman. ÛÏMonitoring Recreational Freshwaters,Û Lakelines, vol 29, pp. 18-24, 2009. [2]åÊ I. Stewart et al.,åÊ ÛÏRecreational and occupational field exposure to freshwater cyanobacteria ÛÒ a review of anecdotal and case reports, epidemiological studies and the challenges for epidemiologic assessmentÛ, Environmental Health: A Global Access Science Source, vol 5, no. 1, 6, 2006. [3] E.D. Hilborn et al., ÛÏ Algal BloomÛÒAssociated Disease Outbreaks Among Users of Freshwater Lakes — United States, 2009ÛÒ2010,ÛåÊ Centers for Disease Control MMWR, vol. 63, no. 1, pp.11-15, Jan. 2014. [4] J. Marshall,åÊ ÛÏIn Summer, Toxic Blue-green Algal Blooms Plague FreshwaterÛ,åÊ Food Environ. Report. Network.åÊ Available:åÊ http://thefern.org/2012/09/in-summer-toxic-blue-green-algae-blooms-plague-freshwater/ [5], L.C. Backer et al., ÛÏCanine Cyanotoxin Poisonings in the United States (1920sÛÒ2012): Review of Suspected and Confirmed Cases from Three Data Sources,Û Toxins, vol 5, pp. 1597-1628, 2013. [6].åÊ P.M. Visser et al., ÛÏThe ecophysicology of the harmful cyanobacterium Microcystis,Û in Harmful Cyanobacteria, J. Huisman, H.C.P. Matthijs and P.M. Visser, eds. Dordrecht, Springer, 2005. [7]åÊ G. A. Codd et al., ÛÏHarmful Cyanobacteria,ÛåÊ in Harmful Cyanobacteria, J. Huisman, H.C.P. Matthijs and P.M. Visser, eds. Dordrecht, Springer, 2005. [8] Chorus, and J. Bartram. Toxic cyanobacteria in water: A guide to their public health consequences, monitoring and management. London, Spon Press, 1999. [9]åÊ T. Kutser et al., “Monitoring cyanobacterial blooms by satellite remote sensing.” Estuarine, Coastal and Shelf Science, vol. 67, no.1, pp. 303-312, 2006. [10]åÊ R.P. Stumpf et al. “Interannual variability of cyanobacterial blooms in Lake Erie.” PloS ONE 7.8,åÊ e42444, 2012. [11]åÊ M.W. Matthews, S. Bernard, L. Robertson, “An algorithm for detecting trophic status (chlorophyll-a) cyanobacterial-dominance, surface scums and floating vegetation in inland and coastal waters.” Remote Sensing of Environment, vol. 124, pp. 637-652, 2012. [12] T.T. Wynne, R. P. Stumpf, and T. O. Briggs. “Comparing MODIS and MERIS spectral shapes for cyanobacterial bloom detection,” Intl Jour. Remote Sensing, vol 34, no. 19, pp.6668-6678. [13]åÊ T.T. Wynne et al.,”Evolution of a cyanobacterial bloom forecast system in western Lake Erie: Development and initial evaluation,” Journal of Great Lakes Research, vol 39, sup 1, pp.90-99. [14]åÊåÊåÊ Ohio EPA,åÊ ÛÏOhio Lake Erie Phosphorus Task Force II Final ReportÛ,åÊ available at http://www.epa.ohio.gov/dsw/lakeerie/index.aspx