Environmental Pollution 240 (2018) 590e598

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Hydrogen peroxide treatment promotes chlorophytes over toxic

cyanobacteria in a hyper-eutrophic aquaculture pond*
Zhen Yang a, b, Riley P. Buley a, Edna G. Fernandez-Figueroa a, Mario U.G. Barros a,
Soorya Rajendran a, Alan E. Wilson a, *
a
Auburn University, School of Fisheries, Aquaculture, and Aquatic Sciences, Auburn, AL 36849, USA
b
State Key Laboratory of Lake Science and Environment, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, 73 East Beijing Road,
Nanjing 210008, China

a r t i c l e i n f o a b s t r a c t

Article history: Controlling blooms of toxigenic phytoplankton, including cyanobacteria, is a high priority for managers
Received 15 December 2017 of aquatic systems that are used for drinking water, recreation, and aquaculture production. Although a
Received in revised form variety of treatment approaches exist, hydrogen peroxide (H2O2) has the potential to be an effective and
2 May 2018
ecofriendly algaecide given that this compound may select against cyanobacteria while not producing
Accepted 3 May 2018
harmful residues. To broadly evaluate the effectiveness of H2O2 on toxigenic phytoplankton, we tested
multiple concentrations of H2O2 on (1) four cyanobacterial cultures, including filamentous Anabaena,
Cylindrospermopsis, and Planktothrix, and unicellular Microcystis, in a 5-day laboratory experiment and
Keywords:
Algaecide
(2) a dense cyanobacterial bloom in a 7-day field experiment conducted in a nutrient-rich aquaculture
Diversity pond. In the laboratory experiment, half-maximal effective concentrations (EC50) were similar for Ana-
Green algae baena, Cylindrospermopsis, and Planktothrix (average EC50 ¼ 0.41 mg L1) but were ~10x lower than
Microcystin observed for Microcystis (EC50 ¼ 5.06 mg L1). Results from a field experiment in an aquaculture pond
Phytoplankton showed that 1.3 and  6.7 mg L1 of H2O2 effectively eliminated Planktothrix and Microcystis, respec-
Zooplankton tively. Moreover, 6.7 mg L1 of H2O2 reduced microcystin and enhanced phytoplankton diversity, while
causing relatively small negative effects on zooplankton abundance. In contrast, 20 mg L1 of H2O2
showed the greatest negative effect on zooplankton. Our results demonstrate that H2O2 can be an
effective, rapid algaecide for controlling toxigenic cyanobacteria when properly dosed.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction systems is necessary for the long-term elimination of harmful

cyanobacterial blooms; however, minimizing runoff inputs and
Eutrophication of freshwater systems is a global predicament managing both sedimentation and internal loading is challenging.
that has important ramifications for the health of aquatic food Nutrient control in aquaculture is especially difficult given the need
webs, animals, and people through the promotion of cyanobacterial to regularly feed farmed fish. Although most water resource man-
blooms, including taxa known to produce toxic secondary metab- agers understand that reducing cyanobacterial blooms should be a
olites such as microcystins and saxitoxins (Neilan et al., 2013; long-term goal, most are keen to find solutions that create quick
Ibelings et al., 2014). Moreover, some cyanobacteria produce taste and noticeable improvements in water quality. Consequently,
and odor compounds, such as geosmin and 2-Methylisoborneol multiple methods have been developed aimed at reducing phyto-
(MIB), that have no known negative human health consequences plankton density or inhibiting their growth, including ultra-
but impart unpleasant musty flavors and odors in drinking water sonication (Ahn et al., 2007a; Lürling et al., 2014), modified clays
and aquaculture products (Zhang et al., 2011; Olsen et al., 2016). (Copetti et al., 2016), and bacterial or chemical agents (Cornish
Undoubtedly, controlling nutrient concentrations in aquatic et al., 2000; Marsalek et al., 2012; Iredale et al., 2012; Greenfield
et al., 2014). Although these techniques can be effective, some
can harm non-target organisms or lead to chemical residual accu-
*
This paper has been recommended for acceptance by Sarah Harmon. mulation in treated ecosystems while others only produce short-
* Corresponding author. lived effects (Matthijs et al., 2016). These drawbacks have
E-mail address: wilson@auburn.edu (A.E. Wilson).

https://doi.org/10.1016/j.envpol.2018.05.012
0269-7491/© 2018 Elsevier Ltd. All rights reserved.
 Z. Yang et al. / Environmental Pollution 240 (2018) 590e598 591

prevented their popularization as control methods for cyano- laboratory experiment study was obtained from the University of
bacterial blooms. Texas at Austin culture collection. Three filamentous cyanobacterial
Hydrogen peroxide (H2O2) is a strong oxidant that is widely cultures, including Anabaena flos-aquae (also called Dolicho-
used for disinfection in water treatment and is on the U.S. Food and spermum flos-aquae) clone R5 (AU pond R5; isolated 15 August
Drug Administration (FDA) approved aquaculture drugs list for 2010), Planktothrix agardhii clone G24 (AU pond E24; isolated 15
aquaculture (U.S. FDA Approved Aquaculture Drugs, assessed 11 August 2010), and Cylindrospermopsis raciborskii clone R2 (AU pond
November 2017). Since H2O2 rapidly decomposes to H2O and O2 R2; isolated 29 August 2010), were originally isolated from aqua-
via biological, chemical, and photochemical mechanisms during culture ponds located at the E. W. Shell Fisheries Center of Auburn
oxidation, it does not leave harmful residues in the environment. Its University in Auburn, Alabama. Axenic batch cultures of each strain
strong oxidizing ability also promotes algal cell mortality by pro- were diluted with 60 ml of BG11 medium to achieve similar initial
ducing hydroxyl radicals under light exposure, which destroys phycocyanin concentrations (filter blank corrected; data reported
proteins, lipids, and DNA (Latifi et al., 2009). More importantly, as raw fluorescence units (RFU); ~3000 RFU) with the CyanoFluor
cyanobacteria are more sensitive than other phototrophs to H2O2 (Turner Designs, CA, USA) and added to 100 ml glass flasks at 25  C
because of their unique cellular structure (Dra bkova  et al., 2007). on a12:12 h light:dark cycle at 40 mmol m2 s1 photosynthetically
Thus, H2O2 is expected to be a selective algaecidal compound to active radiation. Hydrogen peroxide (Baker Analyzed, 30%) was
control cyanobacterial blooms and may be a sensible alternative to added into each flask to achieve the follow concentrations:
controlling cyanobacterial blooms. 0 (control), 0.3, 0.9, 2.7, 8.1, and 24.3 mg L1. Each treatment was
Prior studies have highlighted the potential of H2O2 for con- replicated three times. Photosynthetic changes associated with
trolling cyanobacteria while also promoting non-toxic phyto- H2O2 additions, including photosynthetic activity and phycocyanin
plankton. For example, Barroin and Feuillade (1986) found the pigments, were monitored before the treatments as well as 1, 2, and
toxicity threshold of Oscillatoria (renamed to Planktothrix) under 5 d after H2O2 was added to each flask. Due to sample volume
laboratory conditions to be 1.75 mg L1 H2O2, whereas the domi- constraints, samples were collected from each flask using sterile
nant chlorophyte, Pandorina sp, was unaffected at a 10x higher pipet tips and analyzed for quantum efficiency (measured as Fv/Fm,
H2O2 dose. Moreover, Dra bkova et al. (2007) also showed that where Fv is the maximal variable fluorescence and Fm is the
cyanobacteria were negatively affected by H2O2 at concentrations maximal fluorescence intensity) using the Aquapen C100 (Photon
10 times less than that of green algae and diatoms, and that high Systems Instruments, Brno, Czech Republic) after placing the
light enhanced this effect across phytoplankton taxa. Lastly, samples in the dark for 5 min at room temperature. Flask sub-
Matthijs et al. (2012) performed an unreplicated experiment in a samples were also analyzed for phycocyanin concentration with
mesotrophic lake to test if H2O2 can be used to selectively suppress the CyanoFluor.
cyanobacteria in natural waters without affecting other organisms.
In that study, Planktothrix was reduced by 99% only 3 d after the 2.2. Field experiment
addition of 2 mg L1 of H2O2, while eukaryotic phytoplankton and
zooplankton remained largely unaffected. More recently, a field The mesocosm experiment was conducted in a small (1.5
experiment confirmed that H2O2, used as sodium carbonate per- hectare), shallow (maximum depth ¼ 2.7m), hyper-productive
oxyhydrate (a granular source of H2O2), caused a decline of (~1400 mg L1 total nitrogen and 122 mg L1 total phosphorus) cat-
phycocyanin concentrations and cell densities but did not affect fish aquaculture pond (S9; Boyd and Shelton, 1984) located at the E.
chlorophyll a concentrations (Geer et al., 2017). Clearly, hydrogen W. Shell Fisheries Center of Auburn University during May 2017. At
peroxide has potential for controlling cyanobacteria in diverse the start of the experiment, there was a toxic cyanobacterial bloom
systems. dominated by Planktothrix and Microcystis. Twelve greenhouse
Since freshwater algal blooms can be dominated by one or many plastic limnocorrals (1500 L) that were sealed at the bottom and
cyanobacterial genera, including Anabaena, Aphanizomenon, Cylin- open at the top were suspended from a floating PVC frame in the
drospermopsis, Microcystis, and Planktothrix, it would be useful to pond and filled by pumping pond water through a sieve (500 mm)
know if interspecific variation exists in H2O2 toxicity thresholds to remove small fish. To increase the biomass of Microcystis colonies
across important bloom-forming genera. Past efforts aimed at using in the enclosures, we collected plankton with a 100 mm net and
H2O2 to control cyanobacterial blooms have focused on Planktothrix then added similar volumes of the concentrated plankton homog-
and Microcystis. Few studies exist documenting the effect of H2O2 enate into each enclosure. All enclosures were fertilized with KNO3
on Anabaena, Aphanizomenon, and Cylindrospermopsis. Across these and K2HPO4 to reach the target nutrient levels (4 mg L1 for total
studies, H2O2 toxicity thresholds span several orders of magnitude nitrogen and 0.2 mg L1 for total phosphorus) and sampled at 9:00
from 0.33 to 60 mg L1 (Barrington et al., 2013; Bauza et al., 2014; am on day 0 before establishing the four H2O2 treatments (0, 1.3,
Wang et al., 2012). 6.7, and 20 mg L1 H2O2, single application; three replicates per
In this study, we assessed the toxicity of H2O2 on three fila- treatment). One enclosure was damaged during the experiment
mentous and one currently unicellular but originally colonial cya- and data collected from this enclosure was not used for later ana-
nobacterial genera under laboratory conditions. Based on results lyses. Each enclosure was mixed thoroughly prior to collecting in-
from the laboratory study, we conducted a replicated, field enclo- tegrated samples (surface to 1 m depth) with a rigid tube sampler at
sure experiment in a hyper-eutrophic aquaculture pond to inves- 9:00 am 1, 3, 5, and 7 days after the H2O2 addition. Samples were
tigate the effect of four different H2O2 concentrations on the returned to the lab and processed for phytoplankton and
plankton community, which included a dense cyanobacterial zooplankton diversity and abundance, two algal pigments (chlo-
bloom dominated by toxic, filamentous Planktothrix and colonial rophyll a and phycocyanin), and the hepatotoxin microcystin (both
Microcystis, as well as the associated zooplankton community. intracellular (particulate) and extracellular (dissolved)).
Phytoplankton samples were preserved with 1% Lugol's iodine
2. Materials and methods solution before settling a small volume (180 ml) in a Palmer cham-
ber and enumerating all phytoplankton taxa observed in 10 fields
2.1. Laboratory experiment under 100x magnification (Olsen et al., 2016). Nikon image soft-
ware was used to estimate the biovolume for each taxon.
The unicellular Microcystis aeruginosa (UTEX 2667) used for the Zooplankton were collected using a 60 mm filter and preserved in
592 Z. Yang et al. / Environmental Pollution 240 (2018) 590e598

95% aqueous ethanol. One milliliter of each zooplankton sample

was settled in a Sedgwick Rafter chamber and all copepods, cla-
docera, and rotifers were enumerated under 40x magnification
with a Nikon microscope. Filters (Pall A/E) for chlorophyll a and
phycocyanin concentrations were collected from well-mixed sam-
ples and stored frozen in the dark until being analyzed using a
Turner Designs Trilogy fluorometer with pigment specific modules
(Kasinak et al., 2015). Chlorophyll a samples were measured fluo-
rometrically after a 24 h extraction in 90% aqueous ethanol in
darkness at 4  C (Kasinak et al., 2015). Phycocyanin samples were
measured fluorometrically after grinding, extracting each filter in a
50 mM phosphate buffer in darkness at 4  C, and filtering (<0.2 mm)
each sample (Kasinak et al., 2015). Filters for intracellular micro-
cystin were also collected from well-mixed samples and stored
frozen in the dark until being extracted with 75% acidic aqueous
methanol and analyzed with a BioTek 96-well plate readers using
enzyme-linked immunosorbent assay (ELISA; Chislock et al., 2014).
Sample filtrate (<1 mm) was also analyzed using ELISA for extra-
cellular microcystin.

Fig. 1. The effect of six different concentrations of hydrogen peroxide (0, 0.3, 0.9, 2.7,
8.1, and 24.3 mg L1) on four cyanobacterial cultures (Anabaena, Cylindrospermopsis, 2.3. Statistical analyses
Planktothrix, and Microcystis) under laboratory conditions. Filters show variation in
algal abundance across treatments (50 ml of sample collected on each filter).
The half-maximal effective concentration (EC50; also called the

Fig. 2. The effect of six different concentrations of hydrogen peroxide (0, 0.3, 0.9, 2.7, 8.1, and 24.3 mg L1) on photosynthetic activity (Fv/Fm) of four cyanobacterial cultures,
including (A) Anabaena, (B) Cylindrospermopsis, (C) Planktothrix, and (D) Microcystis under laboratory conditions over five days. Fv is the maximal variable fluorescence and Fm is the
maximal fluorescence intensity. Results are expressed as the mean ± SD (standard deviation).
 Z. Yang et al. / Environmental Pollution 240 (2018) 590e598 593

median inhibitory concentration) was used to evaluate the toxicity the treatments 2.7 mg L1 H2O2 (Fig. 2D). No significant effect in
of H2O2 on four different cyanobacterial taxa. The EC50 value, based Fv/Fm was found for Microcystis treated with H2O2 doses
on the phycocyanin data measured on the 2 days after the H2O2 0.9 mg L1 after 5 days (Fig. 2; P ¼ 0.78). Phycocyanin concen-
addition, was calculated for each cyanobacterium using Probit trations (reported as CyanFluor RFUs) showed similar patterns
analysis. The EC50 values were analyzed by one-way analysis of observed for photosynthetic activity (Fig. 3). For example, the
variance (ANOVA). Differences of the H2O2 treatments in the lab- phycocyanin concentration for each filamentous cyanobacterium
oratory and field experiments were compared using ANOVA fol- decreased from 3000 to near 0 after 1 day of H2O2 treatments
lowed by a Tukey's test. The ShannoneWiener diversity index 0.9 mg L1 and was maintained near 0 for the remainder of the 5-
(Shannon and Weaver, 1949) and SimpsoneDominance index day experiment (Fig. 3AeC). In contrast, only H2O2 doses
(Simpson, 1949) were used to evaluate the effect of H2O2 on 8.1 mg L1 produced the same inhibitory effect in Microcystis over
phytoplankton diversity. All statistical analyses were performed five days (Fig. 3D). The phycocyanin concentration of Microcystis
with SPSS 13.0. treated with 0.9 and 2.7 mg L1 of H2O2 remained at ~17,000 or
~3,000, respectively, despite being lower than the controls
3. Results (~27,000; Fig. 3D) by the end of the experiment. The H2O2 EC50
values were similar for the three filamentous cyanobacteria (~0.41)
The four cyanobacterial genera showed large variation in but an order of magnitude lower than the EC50 observed for
their resistance to H2O2 (Fig. 1). For example, H2O2 doses of Microcystis (5.06; Table 1).
0.9 mg L1 significantly decreased the pigment concentration Observations made during the field experiment were generally
(Fig. 1) and Fv/Fm value in three filamentous cyanobacteria (Fig. 2; similar to those observed for the laboratory-based experiment. For
Anabaena: P < 0.001; Cylindrospermopsis: P < 0.00001; Planktothrix: example, after 1 day, H2O2 addition caused significant decreases in
P < 0.0001) after only 1 day and was maintained at low levels for chlorophyll a (Fig. 4A; P ¼ 0.022) and phycocyanin (Fig. 4B;
each filamentous cyanobacterial taxa until day 5 (P < 0.01). In P < 0.0001) concentrations. However, chlorophyll a concentrations
contrast, a higher H2O2 dose (2.7 mg L1) was needed to signifi- in all treatments recovered and were not different compared with
cantly reduce Fv/Fm for Microcystis after 1 d (P < 0.00001). After 5 d the control after 7 days (Fig. 4A; P ¼ 0.20). For phycocyanin, only the
of the H2O2 addition, photosynthetic activity was only detected in 1.3 mg L1 treatments returned to the same concentration as the

Fig. 3. The effect of six different concentrations of hydrogen peroxide (0, 0.3, 0.9, 2.7, 8.1, and 24.3 mg L1) on phycocyanin concentration (measured as raw fluorescence units (RFU))
of four cyanobacterial cultures, including (A) Anabaena, (B) Cylindrospermopsis, (C) Planktothrix, and (D) Microcystis under laboratory conditions over five days. Results are expressed
as the mean ± SD.
594 Z. Yang et al. / Environmental Pollution 240 (2018) 590e598

Table 1
The half-maximal effective concentration (EC50; mg L1) of hydrogen
peroxide on four cyanobacterial cultures (Anabaena, Cylin-
drospermopsis, Planktothrix, and Microcystis) under laboratory con-
ditions. Results are expressed as the mean ± SD.

Cyanobacterial genera EC50(mg L1)

Anabaena 0.50 (±0.04)a

Cylindrospermopsis 0.32 (±0.02)a
Planktothrix 0.42 (±0.04)a
Microcystis 5.06 (±0.24)
a
Indicate a significant difference with Microcystis (P < 0.01).

Fig. 5. The effect of four different concentrations of hydrogen peroxide (0, 1.3, 6.7, and
20 mg L1) on (A) Planktothrix and (B) Microcystis biovolume (mm3 mL1) during a 7-
day field experiment. Results are expressed as the mean ± SD.

(P ¼ 0.133 compared to control). Higher doses of H2O2 (6.7 and

20 mg L1) caused rapid and sustained reductions of Microcystis
relative to the control (Fig. 5B; P < 0.001).
The effects of H2O2 treatments on microcystin were interesting
and concentration dependent. For example, intracellular micro-
cystin was not affected by the 1.3 mg L1 H2O2 dose (P ¼ 0.27;
compared to control) after 7 days (Fig. 6A) but rapidly declined after
1 day in the 6.7 and 20 mg L1 H2O2 treatments, from 3.74 to
2.99 mg L1 to 0.13 and 0.27 mg L1, respectively. Intracellular
microcystin was maintained at low concentrations until the
completion of the experiment in these two higher H2O2 treatments
(Fig. 6A; P  0.058 compared to control). Extracellular microcystin
increased quickly in all H2O2 treatments 1 day after treatment due
to the release of intracellular toxins, but then significantly declined
to low concentrations similar to those in the control enclosures by
day 7 (Fig. 6B; P > 0.06).
Fig. 4. The effect of four different concentrations of hydrogen peroxide (0, 1.3, 6.7, and Plankton diversity was affected by H2O2 treatments. For
20 mg L1) on (A) chlorophyll a and (B) phycocyanin concentrations (mg L1) during a
instance, 6.7 and 20 mg L1 of H2O2 enhanced phytoplankton di-
7-day field experiment. Results are expressed as the mean ± SD.
versity (Table 2). Compared with their initial values, 6.7 and
20 mg L1 treatments more than doubled the ShannoneWiener
diversity index after 7 days, while a significant diversity increase
control (Fig. 4B; P ¼ 0.981) after 7 days, while the 6.7 and 20 mg L1 was also observed for the SimpsoneDominance index for
addition enclosures remained at very low concentrations (Fig. 4B) 6.7 mg L1 treatment. Regarding zooplankton, repeated measure
throughout the experiment. analysis showed that the 1.3 and 6.7 mg L1 H2O2 treatment led to
Planktothrix and Microcystis dominated all of the enclosures at similar total zooplankton density (including copepods, cladocera,
the start of the experiment (Fig. 5; Table 2). The diameter of and rotifers) as the control (P > 0.05; Fig. 7A). Significant inhibitory
Microcystis colonies in our enclosures ranged from 16 to 280 mm effects of H2O2 on zooplankton were observed in the enclosures
(72 mm mean diameter). Of these two cyanobacterial taxa, the treated with 20 mg L1 (P < 0.05; Fig. 7A).
lowest H2O2 dose (1.3 mg L1) proved to be most effective against
Planktothrix (Fig. 5A). Planktothrix biovolume gradually decreased
from 1.03 - 1.39  108 mm3 mL1 to 0.04 - 0.08  108 mm3 mL1 in all 4. Discussion
treated enclosures after 7 days (Fig. 5A). In contrast, Microcystis
showed a higher resistance to the 1.3 mg L1 H2O2 dose after 7 days Hydrogen peroxide was shown to be an effective treatment
 Z. Yang et al. / Environmental Pollution 240 (2018) 590e598 595

Table 2
The effect of four different concentrations (0, 1.3, 6.7, and 20 mg L1) of hydrogen peroxide on phytoplankton taxa biovolume (mm3 mL1) and two phytoplankton diversity
indices (Shannon-Weiner Diversity and Simpson-Dominance Index) at the start (day 0) and end (day 7) of the field experiment.

Day 0 Day 7

0 mg L1 1.3 mg L1 6.7 mg L1 20 mg L1 0 mg L1 1.3 mg L1 6.7 mg L1 20 mg L1

Cyanobacteria 108,744,412 121,956,891 102,218,486 109,717,503 202,797,641 255,822,171 16,361,364 1,510,184

Anabaena 1,298,605 1,056,328 1,347,061 1,763,777 3,130,220
Arthrospira 160,687 160,687
Chroococcus 124,210 180,014 125,967 183,615 831,667 2,603,009 13,236,464 1,071,986
Microcystis 96,658,914 106,752,104 90,300,590 96,703,365 190,669,103 252,396,979 2,704,346
Planktothrix 10,501,996 13,968,444 10,284,181 11,066,746 8,166,651 822,183 420,555 438,198

Chlorophytes 1,612,878 2,427,864 4,535,885 6,961,571 924,919 1,610,040 8,820,604 8,514,181

Actinastrum 164,991 54,997
Ankistrodesmus 1105 13,256 2209 3314 1105 134,885 1657
Chlamydomonas 15,444 57,916 19,305 81,082 19,305 37,779 23,166
Chlorella 2919 2919 6567 6567 164,166 299,148
Closterium 42,730 3662 119,330 1831 2136 39,373
Coelastrum 1,141,463 1,712,194 3,424,389 6,278,046 88,261 220,652 2,471,302 926,738
Cosmarium 75,784
Golenkinia 26,045 26,045 31,275 23,457 109,464 31,275 838,672 6,016,406
Oocystis 56,850 48,095 46,129 7813 7813 24,562 186,599
Pandorina 186,690 258,929
Pediastrum 165,438 496,314 661,752 496,314 496,314 330,876 2,481,569 496,314
Scenedesmus 29,319 26,308 43,847 13,154 32,075 122,771 1,025,063 604,373
Schroederia 11,678 7785
Selenastrum 14,784
Staurastrum 188,416 156,376 168,446 544,755 1,128,211
Tetraedron 24,304
Treubaria 8097 22,740 22,740

Diatoms 17,114 3377 128,187 13,809 101,522 0 48,461 0

Aulacoseira 91,506
Cyclotella 101,522
Fragilaria 22,806
Navicula 17,114 3377 36,681 13,809 1126
Synedra 24,529

Others 1,259,504 1,816,323 2,346,955 1,029,900 1,241,834 7,818,366 746,660 2,705,722

Euglena 96,475 192,950 199,542 144,712 482,374 192,950 806,017
Peridiniopsis 283,341 366,676 258,257 125,003 83,336 6,928,361 565,089 650,017
Trachelomonas 571,386
Cryptomonas 879,688 1,256,697 1,889,157 760,185 676,124 125,670 181,570 1,249,688

ShannoneWiener Diversity Index 0.54 0.60 0.75 0.69 0.37 0.29 1.77 1.85

SimpsoneDominance Index 0.23 0.17 0.35 0.35 0.14 0.10 0.59 0.25

against toxigenic cyanobacteria in both laboratory- and field-based been studied individually (Dr abkova et al., 2007; Matthijs et al.,
experiments. In the latter experiment, high quality and harmless 2012; Wang et al., 2012; Bauza et al., 2014), this is the first time
chlorophytes were promoted by higher H2O2 treatments H2O2 toxicity has been tested on both Anabaena and Cylin-
(>6.7 mg L1), which should improve trophic transfer efficiency in drospermopsis concurrently with Planktothrix and Microcystis to
productive aquatic systems. The toxicity of H2O2 to phytoplankton compare its toxicity under similar conditions. Due to the high
is mainly attributed to the production of hydroxyl radicals that morphological diversity of cyanobacterial species included in our
destroy cell membrane integrity (Mikula et al., 2012) and may lab experiment, it was challenging to reach an equal initial biomass
damage the photosynthetic apparatus thus inhibiting photosyn- for each species by cell number or volume. Phycocyanin concen-
thetic electron transfer (Barrington and Ghadouani, 2008). Higher tration has been generally used to estimate the cyanobacteria
sensitivity of cyanobacteria to H2O2 than other associated phyto- biomass in the field as it correlates to cyanobacterial biomass and is
plankton was observed more than three decades ago (Barroin and easy to measure (Ahn et al., 2007b; Chang et al., 2012; Bowling
Feuillade, 1986) and is likely due to the inability of cyanobacteria et al., 2016). For this reason, the same phycocyanin concentration
to eliminate reactive oxygen (Latifi et al., 2009) or that their was used to represent the approximately equal initial biomass
photosynthetic apparatus is connected directly to the cellular levels for each species.
plasma membrane (Grossman et al., 1995). These mechanisms Microcystis showed a remarkably higher (10x) resistance to
could explain our laboratory-based observations with four cultures H2O2 than the filamentous cyanobacterial taxa tested both in our
of cyanobacteria that showed a rapid decrease in the quantum ef- laboratory and field experiments (Figs. 1e3, 5; Table 1). Some
ficiency (measured as Fv/Fm; Fig. 2) of the cultures immediately mechanisms that have been used to explain the competitive
after the addition of H2O2. Moreover, the rapid release of micro- dominance of Microcystis in diverse freshwater systems include its
cystin in the two highest H2O2 treatments in the field experiments higher floating regulation, colonial morphology, and high-
(Fig. 6B) suggests a H2O2-mediated disruption in the cyanobacterial irradiance resistance (Huisman et al., 2005). However, these
cell membrane. However, cell membrane integrity was not evalu- mechanisms cannot explain higher resistance of Microcystis to H2O2
ated in this study. in this study since the experiment was conducted in the laboratory
Although the toxicity of H2O2 to Planktothrix and Microcystis has with relatively low light compared to nature and that the
596 Z. Yang et al. / Environmental Pollution 240 (2018) 590e598

(Fig. 4B) while other phytoplankton, especially chlorophytes, quickly

rebounded. Green algae biomass did not increase significantly in the
1.3 mg L1 H2O2 treatment because Microcystis was not affected by
the low dose of H2O2. These results are consistent with prior field
experiments (Matthijs et al., 2012; Weenink et al., 2015). Although
6.7 mg L1 is near the H2O2 toxicity threshold of specific non-
cyanobacterial phytoplankton (Dra bkova et al., 2007), we contend
that the reduction of competition from bloom-forming cyanobac-
teria creates a better habitat for other, harmless phytoplankton taxa
thus increasing their abundance and diversity. Indeed, the
enhancement of phytoplankton diversity after H2O2 treatment is
also supported by the increase of both ShannoneWiener diversity
and SimpsoneDominance indices (Table 2). However, the 20 mg L1
H2O2 treatment showed harmful effects to some non-cyanobacterial
phytoplankton taxa, which reduced species numbers and conse-
quently decreased the SimpsoneDominance index despite an in-
crease in the ShannoneWiener diversity index (Table 2).
Since H2O2 treatments could also impact zooplankton, which
may influence phytoplankton species composition and abundance
through grazing and nutrient recycling as well as fish production
through planktivory, we were also interested in the effects of
different H2O2 doses on the zooplankton community. Matthijs et al.
(2012) found that the addition of 2 mg L1 H2O2 did not cause
harmful effect on zooplankton. In this field experiment, the two
lower H2O2 treatments (0 and 1.3 mg L1) significantly increased
total zooplankton biomass (especially rotifers and cladocera; Fig. 7)
in response to small reductions in cyanobacteria (Fig. 4) or intra-
cellular microcystin (Fig. 6). Total zooplankton density was
Fig. 6. The effect of four different concentrations of hydrogen peroxide (0, 1.3, 6.7, and marginally affected by the 6.7 mg L1H2O2 treatment. However, the
20 mg L1) on (A) intracellular and (B) extracellular microcystin concentrations (mg highest H2O2 treatment (20 mg L1) quickly caused significant de-
L1) during a 7-day field experiment. Results are expressed as the mean ± SD.
clines in total zooplankton abundance relative to the control that
persisted during the 7-day field experiment (Fig. 7A). Thus, water
Microcystis culture used was unicellular. In addition, Dziallas and resource managers should consider dosing needs when controlling
Grossart (2011) found that H2O2 toxicity to toxic Microcystis cyanobacterial blooms given that higher doses could have negative
strains is significantly lower than non-toxic conspecifics because consequences on non-target organisms, such as zooplankton.
microcystins may play a role in stabilizing the photosynthetic Although the toxicity of H2O2 to other filamentous cyano-
apparatus and enhances the fitness of toxin producing strains un- bacterial taxa (except for Planktothrix) was not evaluated in the field
der oxidative stress (Phelan and Downing, 2011; Zilliges et al., experiment, H2O2 may be an effective algaecide to control other
2011). Whether differences in the ability to produce toxic second- filamentous cyanobacteria based on the results from this and past
ary metabolites, like microcystin, of the cyanobacterial taxa used in studies (Matthijs et al., 2012; Bauza et al., 2014). In these studies,
this study is responsible for the variation in observed resistance to lower H2O2 concentrations (e.g., ~1.3 mg L1) were sufficient to
H2O2 is in need of further study. eliminate most filamentous cyanobacterial cells without large
Intracellular microcystin concentrations are dependent on the negative effects on non-target taxa. However, Microcystis appears
biomass of cyanobacterial cells as well as the internal toxin pro- to have a higher H2O2 toxicity threshold (Table 1). The 6.7 mg L1
duction and storage per unit cell. Consequently, in the field H2O2 treatment showed to be effective at quickly eliminating
experiment, we found large (>90%) decreases in intracellular Microcystis colonies and microcystin during the enclosure experi-
microcystin concentration in the two highest H2O2 treatments (6.7 ment, while also having small negative effects on zooplankton
and 20 mg L1; Fig. 6A) concomitant with large reductions (90%) in density and positive effects on phytoplankton diversity. A sug-
cyanobacterial abundance relative to 0 (control) and 1.3 mg L1 gested dose of 6.7 mg L1 H2O2 is lower than similar experiments
H2O2 treatments (Fig. 4, Table 2). Toxin release after cell disruption performed by Wang et al. (2012). One reason for this discrepancy
is a key problem associated with the use of algaecides in the control could be due to Microcystis colonies in their experiment (diameter
of cyanobacteria. In the field experiment, extracellular microcystin above 100 mm) being larger than those found in our enclosures
concentration rose rapidly after the addition of H2O2 in all treat- (72 mm mean colony diameter). Previous studies have shown that
ments (Fig. 5B). However, H2O2 can enhance the photocatalytic bigger colony size increases tolerance of Microcystis to H2O2 toxicity
oxidation of microcystins, making them degrade into non-toxic by- (Liu et al., 2017). Another explanation for this discrepancy could be
products (Rodriguez et al., 2007; Lürling et al., 2014). Consistent that the Microcystis colony density in this study was not as high as
with previous studies (Matthijs et al., 2012; Papadimitriou et al., that of Wang et al. (2012). It may be that a dose higher than
2016), the results of our field experiment study show that H2O2 6.7 mg L1 H2O2 is needed to control Microcystis blooms with more
degraded extracellular toxins and significantly reduced their levels dense or larger colonies. Considering the non-target effects of
by the end of the experiment (Fig. 6B). higher H2O2 doses on zooplankton, care is encouraged during
Understanding the effect of H2O2 on non-target organisms treatment dosing. However, an appropriate dose of H2O2, such as
should be assessed before its use in any aquatic ecosystem and 6.7 mg L1, should be considered for use in waterbodies with low
especially commercial aquaculture ponds. Although phytoplankton Microcystis biomass or in the early stage of a Microcystis bloom
were generally reduced after H2O2 additions (Fig. 4A), cyanobacteria dominated by small colonies (Zhu et al., 2016).
were consistently repressed throughout the 7-day field experiment Ultraviolet radiation and high direct sunlight conditions could
 Z. Yang et al. / Environmental Pollution 240 (2018) 590e598 597

Fig. 7. The effect of four different concentrations of hydrogen peroxide (0, 1.3, 6.7, and 20 mg L1) on zooplankton density (individuals L1) during a 7-day field experiment. (A) Total
zooplankton (Copepods þ Cladocera þ Rotifers), (B) Copepods, (C) Cladocera, and (D) Rotifers. Results are expressed as the mean ± SD.

promote the reaction of H2O2 and produce more hydroxyl radicals, also promoting other phytoplankton taxa, including high quality
a strong reactive oxygen species that can damage cyanobacterial chlorophytes. Negative effects on zooplankton at the highest H2O2
cells by destroying their membrane integrity and photosynthetic concentration (20 mg L1) highlight the need to conduct small-
apparatus (Mikula et al., 2012; Barrington and Ghadouani, 2008). scale studies testing various H2O2 concentrations prior to whole
According to results in Dr
abkova  et al. (2007), initial concentrations system treatments.
of H2O2 were decomposed to less than 4% after 3 h of exposure
under high irradiance, but showed greater toxicity to algal cells Acknowledgments
than under the lower irradiance. Though H2O2 concentrations were
not measured in this experiment, the H2O2 levels were expected to This project was supported by the National Natural Science
decrease quickly over time based on two large-scale field H2O2 Foundation of China (grant # 31570457), the United States
manipulations (Matthijs et al., 2012; Burson et al., 2014). In these Department of Agriculture’s National Institute of Food and Agri-
studies, H2O2 declined by  60% or 95% within one or two days culture (grant # 2017-70007-27132), and the School of Fisheries,
after treatment, respectively. Consequently, we contend that for Aquaculture, and Aquatic Sciences and the College of Agriculture at
H2O2 to affect algal cells prior to breaking down, the treatment Auburn University.
must be evenly, quickly, and thoroughly mixed. Future research
should evaluate the effects of different H2O2 application concen- Appendix A. Supplementary data
trations under different light intensities and UV radiation.
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.envpol.2018.05.012.
5. Conclusions
References
Blooms of cyanobacteria pose serious threats to aquatic eco-
systems around the world. Consequently, a variety of techniques to Ahn, C.Y., Joung, S.H., Choi, A., Kim, H.S., Jang, K.Y., Oh, H.M., 2007a. Selective control
of cyanobacteria in eutrophic pond by a combined device of ultrasonication and
control cyanobacterial blooms have been developed and tested. We
water pumps. Environ. Technol. 28, 371e379.
were interested in further testing the utility of H2O2 in a hyper- Ahn, C.Y., Joung, S.H., Yoon, S.K., Oh, H.M., 2007b. Alternative alert system for
eutrophic aquaculture pond that regularly suffers from toxic cya- cyanobacterial bloom, using phycocyanin as a level determinant. J. Microbiol.
nobacterial blooms. Using a gradient design field experiment, we 45, 98e104.
Barrington, D.J., Ghadouani, A., 2008. Application of hydrogen peroxide for the
found that H2O2 doses >1.3 mg L1 where effective at significantly removal of toxic cyanobacteria and other phytoplankton from wastewater.
reducing cyanobacteria and the hepatotoxin, microcystin, while Environ. Sci. Technol. 42, 8916e8921.
598 Z. Yang et al. / Environmental Pollution 240 (2018) 590e598

Barrington, D.J., Reichwaldt, E.S., Ghadouani, A., 2013. The use of hydrogen peroxide Water Res. 46, 6095e6103.
to remove cyanobacteria and microcystins from waste stabilization ponds and Kasinak, J.M.E., Holt, B.M., Chislock, M.F., Wilson, A.E., 2015. Benchtop fluorometry
hypereutrophic systems. Ecol. Eng. 50, 86e94. of phycocyanin as a rapid approach for estimating cyanobacterial biovolume.
Barroin, G., Feuillade, M., 1986. Hydrogen peroxide as a potential algicide for J. Plankton Res. 37, 248e257.
Oscillatoria rubescens D.C. Water Res. 20, 619e623. Latifi, A., Ruiz, M., Zhang, C.C., 2009. Oxidative stress in cyanobacteria. FEMS
Bauza, L., Aguilera, A., Echenique, R., Andrinolo, D., Giannuzzi, L., 2014. Application Microbiol. Rev. 33, 258e278.
of hydrogen peroxide to the control of eutrophic lake systems in laboratory Liu, M.X., Shi, X.L., Chen, C., Yu, L., Sun, C., 2017. Responses of Microcystis colonies of
assays. Toxins 6, 2657e2675. different sizes to hydrogen peroxide stress. Toxins 9.
Bowling, L.C., Zamyadi, A., Henderson, R.K., 2016. Assessment of in situ fluorometry Lürling, M., Meng, D.B., Faassen, E.J., 2014. Effects of hydrogen peroxide and ultra-
to measure cyanobacterial presence in water bodies with diverse cyanobacterial sound on biomass reduction and toxin release in the cyanobacterium, Micro-
populations. Water Res. 105, 22e33. cystis aeruginosa. Toxins 66, 3260e3280.
Boyd, C.E., Shelton, J.L., 1984. Observations on the hydrology and morphometry of Marsalek, B., Jancula, D., Marsalkova, E., Mashlan, M., Safarova, K., Tucek, J.,
ponds on the Auburn university Fisheries research unit. Ala. Agric. Exp. Stat. Zboril, R., 2012. Multimodal action and selective toxicity of zerovalent iron
Bull. 558. nanoparticles against cyanobacteria. Environ. Sci. Technol. 46, 2316e2323.
Burson, A., Matthijs, H.C.P., de Bruijne, W., Talens, R., Hoogenboom, R., Gerssen, A., Matthijs, H.C.P., Jancula, D., Visser, P.M., Marsalek, B., 2016. Existing and emerging
Visser, P.M., Stomp, M., Steur, K., van Scheppingen, Y., Huisman, J., 2014. cyanocidal compounds: new perspectives for cyanobacterial bloom mitigation.
Termination of a toxic Alexandrium bloom with hydrogen peroxide. Harmful Aquat. Ecol. 50, 443e460.
Algae 31, 125e135. Matthijs, H.C.P., Visser, P.M., Reeze, B., Meeuse, J., Slot, P.C., Wijn, G., Talens, R.,
Chang, D.W., Hobson, P., Burch, M., Lin, T.F., 2012. Measurement of cyanobacteria Huisman, J., 2012. Selective suppression of harmful cyanobacteria in an entire
using in-vivo fluoroscopy - effect of cyanobacterial species, pigments, and lake with hydrogen peroxide. Water Res. 46, 1460e1472.
colonies. Water Res. 46, 5037e5048. Mikula, P., Zezulka, S., Jancula, D., Marsalek, B., 2012. Metabolic activity and
Chislock, M.F., Sharp, K.L., Wilson, A.E., 2014. Cylindrospermopsis raciborskii domi- membrane integrity changes in Microcystis aeruginosa - new findings on
nates under very low and high nitrogen-to-phosphorus ratios. Water Res. 49, hydrogen peroxide toxicity in cyanobacteria. Eur. J. Phycol. 47, 195e206.
207e214. Neilan, B.A., Pearson, L.A., Muenchhoff, J., Moffitt, M.C., Dittmann, E., 2013. Envi-
Copetti, D., Finsterle, K., Marziali, L., Stefani, F., Tartari, G., Douglas, G., Reitzel, K., ronmental conditions that influence toxin biosynthesis in cyanobacteria. En-
Spears, B.M., Winfield, I.J., Crosa, G., D'Haese, P., Yasseri, S., Lürling, M., 2016. viron. Microbiol. 15, 1239e1253.
Eutrophication management in surface waters using lanthanum modified Olsen, B.K., Chislock, M.F., Wilson, A.E., 2016. Eutrophication mediates a common
bentonite: a review. Water Res. 97, 162e174. off-flavor compound, 2-methylisoborneol, in a drinking water reservoir. Water
Cornish, B.J.P.A., Lawton, L.A., Robertson, P.K.J., 2000. Hydrogen peroxide enhanced Res. 92, 228e234.
photocatalytic oxidation of microcystin-LR using titanium dioxide. Appl. Catal. Papadimitriou, T., Kormas, K., Dionysiou, D.D., Laspidou, C., 2016. Using H2O2
B Environ. 25, 59e67. treatments for the degradation of cyanobacteria and microcystins in a shallow
Drabkova, M., Admiraal, W., Marsalek, B., 2007. Combined exposure to hydrogen hypertrophic reservoir. Environ. Sci. Pollut. Res. 23, 21523e21535.
peroxide and light - selective effects on cyanobacteria, green algae, and di- Phelan, R.R., Downing, T.G., 2011. A growth advantage for microcystin production by
atoms. Environ. Sci. Technol. 41, 309e314. Microcystis PCC 7806 under high light. J. Phycol. 47, 1241e1246.
Dziallas, C., Grossart, H.P., 2011. Increasing oxygen radicals and water temperature Rodriguez, E., Majado, M.E., Meriluoto, J., Acero, J.L., 2007. Oxidation of microcystins
select for toxic Microcystis sp. PLoS One 6. by permanganate: reaction kinetics and implications for water treatment.
Food and Drug Administration of the United States of America. Approved Aqua- Water Res. 41, 102e110.
culture Drugs. Assessed 11 November 2017. https://www.fda.gov/ Shannon, C.E., Weaver, 1949. The Mathematical Theory of Communication. Uni-
AnimalVeterinary/DevelopmentApprovalProcess/Aquaculture/ucm132954.htm. versity of Ilinois Press, Urbana,IL.
Geer, T.D., Calomeni, A.J., Kinley, C.M., Iwinski, K.J., Rodgers, J.H., 2017. Predicting in Simpson, E.H., 1949. Measurement of diversity. Nature 163, 688.
situ responses of taste- and odor-producing algae in a southeastern US reser- Wang, Z.C., Li, D.H., Qin, H.J., Li, Y.X., 2012. An integrated method for removal of
voir to a sodium carbonate peroxyhydrate algaecide using a laboratory harmful cyanobacterial blooms in eutrophic lakes. Environ. Pollut. 160, 34e41.
exposure-response model. Water Air Soil Poll. 228. Weenink, E.F.J., Luimstra, V.M., Schuurmans, J.M., Van Herk, M.J., Visser, P.M.,
Greenfield, D.I., Duquette, A., Goodson, A., Keppler, C.J., Williams, S.H., Brock, L.M., Matthijs, H.C.P., 2015. Combatting cyanobacteria with hydrogen peroxide: a
Stackley, K.D., White, D., Wilde, S.B., 2014. The effects of three chemical al- laboratory study on the consequences for phytoplankton community and di-
gaecides on cell numbers and toxin content of the cyanobacteria Microcystis versity. Front. Microbiol. 6.
aeruginosa and Anabaenopsis sp. Environ. Manag. 54, 1110e1120. Zhang, X.J., Chen, C., Lin, P.F., Hou, A.X., Niu, Z.B., Wang, J., 2011. Emergency drinking
Grossman, A.R., Bhaya, D., Apt, K.E., Kehoe, D.M., 1995. Light-harvesting complexes water treatment during source water pollution accidents in China: origin
in oxygenic photosynthesis: diversity, control, and evolution. Annu. Rev. Genet. analysis, framework and technologies. Environ. Sci. Technol. 45, 161e167.
29, 231e288. Zhu, W., Zhou, X.H., Chen, H.M., Gao, L., Xiao, M., Li, M., 2016. High nutrient con-
Huisman, J., Matthijs, H.C.P., Visser, P.M. (Eds.), 2005. Harmful Cyanobacteria. centration and temperature alleviated formation of large colonies of Micro-
Springer, Dordrecht. cystis: evidence from field investigations and laboratory experiments. Water
Ibelings, B.W., Backer, L.C., Kardinaal, W.E.A., Chorus, I., 2014. Current approaches to Res. 101, 167e175.
cyanotoxin risk assessment and risk management around the globe. Harmful Zilliges, Y., Kehr, J.C., Meissner, S., Ishida, K., Mikkat, S., Hagemann, M., Kaplan, A.,
Algae 40, 63e74. Borner, T., Dittmann, E., 2011. The cyanobacterial hepatotoxin microcystin binds
Iredale, R.S., McDonald, A.T., Adams, D.G., 2012. A series of experiments aimed at to proteins and increases the fitness of Microcystis under oxidative stress con-
clarifying the mode of action of barley straw in cyanobacterial growth control. ditions. PLoS One 6 (3), e17615.