Cn

SDJa

b

c

d

e

f

h

••••

a

ARRAA

KACHM

1

n[

(

h0

Journal of Hazardous Materials 298 (2015) 178–187

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journa l homepage: www.e lsev ier .com/ locate / jhazmat

an the algicidal material Ca-aminoclay be harmful when applied to aatural ecosystem? An assessment using microcosms

eung Won Jung a,∗, Suk Min Yun b, Jae Won Yoo c, Zhun Li a,d, Pung-Guk Jang a,hong-Il Lim a,∗∗, Young-Chul Lee e, Hyun Uk Lee f, Taek-Kyun Lee a, Jinbee Heo a,

in Hwan Lee b, Myung-Soo Han d

South Sea Institute, Korea Institute of Ocean Science & Technology, Geoje 656-830, Republic of KoreaDepartment of Life Science, Sangmyung University, Seoul 110-743, Republic of KoreaKorea Institute of Coastal Ecology, Incoperation, Bucheon 421-742, Republic of KoreaDepartment of Life Science, Hanyang University, Seoul 133-791, Republic of KoreaDepartment of BioNano Technology, Gachon University, Seongnam 461-701, Republic of KoreaAdvanced Nano-Surface Research Group, Korea Basic Science Institute, Daejeon 305-333, Republic of Korea

i g h l i g h t s

The effects of using Ca-aminoclay to treat harmful algal blooms need to be understood.Ca-aminoclay was found to be an effective and species-specific algicide.Ca-aminoclay caused bacterioplankton abundances to rapidly increase.Ca-aminoclay had a severe adverse impact on the environment.

r t i c l e i n f o

rticle history:eceived 19 January 2015eceived in revised form 4 May 2015ccepted 10 May 2015vailable online 14 May 2015

eywords:lgicidal materiala-aminoclay

a b s t r a c t

We assessed the ability of an artificial clay (Ca-aminoclay) to suppress harmful algal bloom species (HABs)such as Cochlodinium polykrikoides and Chattonella marina and investigated the ecological responses inthe closed and open microcosm systems. The Ca-aminoclay induced rapidly and selectively cell lysisin the HABs. However, applying Ca-aminoclay could cause adverse impacts in terms of biological andenvironmental changes. The bacterioplankton abundance increased and then, the abundances of het-erotrophic nanoflagellates and ciliates increased rapidly. Extremely poor environmental conditions suchas increase in nutrients and development of anoxic conditions were sustained continuously in a closedsystem, while the environmental conditions in open systems deteriorated before recovering to the initial

armful algal bloomsicrocosm experiments

conditions. We evaluated the potential for the occurrence of a bloom of another phytoplankton afterHABs had been controlled using the Ca-aminoclay. The Ca-aminoclay controlled blooms of Chattonellamarina in mixed cell cultures containing a Tetraselmis chui. However, T. chui increased over time and thenbloomed. Therefore, caution should be taken when considering the direct application of Ca-aminoclay innatural environments even though it offers the rapid removal of HABs.

. Introduction

The occurrence of harmful algal blooms (HABs) is not a new phe-omenon; they are recorded in the Bible and in the fossil record1]. HABs have had major impacts on various marine organisms

∗ Corresponding author. Tel.: +82 556398430.∗∗ Corresponding author. Tel.: +82 556398580.

E-mail addresses: diatoms@kiostac.kr (S.W. Jung), oceanlim@kiost.ac.krD.-I. Lim).

ttp://dx.doi.org/10.1016/j.jhazmat.2015.05.012304-3894/© 2015 Elsevier B.V. All rights reserved.

© 2015 Elsevier B.V. All rights reserved.

throughout history. In particular, HABs often cause the catastrophicmass mortality of fish in near-shore fish farms [2,3]. The regularoccurrence of HABs causes losses of these magnitudes in manycoastal areas around the world [4]. Smayda [5] argued that thereshould be some debate about the nature of HABs and the causesof their recent proliferation, which could be classed as a globalepizootic that is linked to pollution and anthropogenic changes

in coastal ecosystems. Until recently, many researchers consid-ered the best solution to the problem of HABs was to control thealgae themselves. The strategies currently available to inhibit thegrowth of HABs include chemical and biological treatments. The

rdous

mTaihearmsdtteea

tfetacuhtetowtetbhadrceAb(

2

2

dShacoWtwcsi

taC

S.W. Jung et al. / Journal of Haza

ost common at present is the direct application of chemicals.his approach has serious drawbacks, in that species of organismre killed indiscriminately, and blooms recur [6]. Other strategiesnclude bio-manipulation to control the HABs. Bio-manipulationas several disadvantages such as unstable conditions under realnvironments for strong grazing pressure and mass culturing of

predator [7,8]. Therefore, before a new algicidal agent can beeleased into a natural system its efficiency and ecological safety

ust be assessed. Classically, the effects have been evaluated usingingle species toxicity tests in the laboratory [9]. Single species testso not directly provide information on the potential indirect effectshat may occur in a complex ecosystem [10]. The accuracy of predic-ions based on such tests depends on the reliability of the exposurestimates and the validity of extrapolating the data from laboratoryffects to the “real world”. Many of the remaining questions can beddressed using multispecies ecosystem tests.

A promising approach to give high removal efficiencies for thereatment of HABs is via a form of mechanical treatment using eco-riendly non-toxic natural clay [11,12]. However, these treatmentsntail the use of massive amounts of clay, and have not been foundo reduce fish deaths. In particular, the treatment of clay has beenssociated with disturbances in benthic communities [13]. To over-ome disadvantages of natural clay, clay-mimicking materials withnique properties have recently been developed [14–16]. Theseave some of the desirable algicidal characteristics of Ca-aminoclayhat make them useful for treating HABs but have no detrimentalffects on the non-harmful phytoplankton community. However,he study performed by Lee et al. [16] did not yield informationn how the Ca-aminoclay could be realistically used because itas a laboratory study. In the study presented here, we attempted

o answer two questions by conducting three indoor microcosmxperiments. The first question was: can the algicidal material con-rol HABs at different flow rates and prevent the reoccurrence oflooms of the same or another species? The second question was:ow do the biotic and abiotic factors change when the material isdded under different conditions? We used 100 L microcosms toetermine how applying Ca-aminoclay can be host-specific in theeal world and how the treatment affects other marine planktonicommunities because of the acclimatization or outbreak of differ-nt species to the environmental changes caused by the treatment.lso, microcosms were designed to evaluate the Ca-aminoclay inoth a closed (without any seawater added) and open ecosystemsseawater added).

. Materials and methods

.1. Microcosm experiments

The synthesis and characteristics of the Ca-aminoclay areescribed in Section 1 in the Supplementary information (SI, Figs.1–S5). The isolation and culture of HAB species and the non-armful phytoplankton community is described in SI Section 2. As

result of a preliminary test to know the optimal inoculation con-entration of Ca-aminoclay, it was revealed that the concentrationf inoculated Ca-aminoclay was 250 mg L−1. (SI Section 3, Fig. S6).e conducted three experiments (Table 1 and Fig. S7, SI Section 4)

o evaluate fluctuations in the biological and environmental factorshen the Ca-aminoclay was introduced. These were (1) a perfectly

losed system (PCS) without any seawater added, (2) a slow inflowystem (SIS) with seawater added at a rate of 60 L d−1, and (3) a fastnflow system (FIS) with seawater added at a rate of 120 L d−1.

The experiment of a PCS included six enclosures (Table 1);wo of these enclosures contained Cochlodinium polykrikoides atn initial concentration of 4.92 × 103 cells mL−1, two containedhattonella marina at the concentration of 1.07 × 104 cells mL−1,

Materials 298 (2015) 178–187 179

and two contained the non-harmful phytoplankton community ata concentration of 2.97 × 103 cells mL−1. One of each of the enclo-sures containing C. polykrikoides (treatment group TGCp), C. marina(treatment group TGCm), and non-harmful phytoplankton commu-nity (treatment group TGPC) was inoculated with the Ca-aminoclayat a concentration of 250 mg L−1. The other microcosms were usedas controls (the C. polykrikoides control group was called CGCp, theC. marina control group was called CGCm, and the non-harmful phy-toplankton community control group was called CGPC), to whichno Ca-aminoclay was added. The experiment of a SIS includedfour enclosures, each containing approximately 80 L of seawater(Table 1). Two of the enclosures contained C. marina and Tetraselmischui (nano sized green alga) in a mixed culture at a concentrationof 2.17 × 104 cells mL−1 and 0.78 × 104 cells mL−1, respectively,and two contained the natural phytoplankton community (notincluding HABs such as pseudo-nitzschia pungens) at a concentra-tion of 2.76 × 104 cells mL−1. The Ca-aminoclay at a concentrationof 250 mg L−1 was inoculated only into the enclosures containing C.marina. Filtered and sterilized natural seawater was supplied to theSIS experiment enclosures at a rate of 60 L d−1. The water contain-ing Ca-aminoclay that overflowed from the C. marina treatmentgroup (TGCm) was fed into one of the enclosures containing thenon-harmful phytoplankton community (TGPC; Ca-aminoclay wasnot inoculated directly in TGPC). The FIS experiment was performedto evaluate resilience of ecosystem (Table 1). Two enclosures eachcontaining approximately 80 L of seawater contained C. marina ata concentration of 1.97 × 104 cells mL−1 and contained the non-harmful phytoplankton community at a concentration of 1.40 × 102

cells mL−1. The FIS experiment enclosures were all inoculated withthe Ca-aminoclay at a concentration of 250 mg L−1. The seawaterwas supplied to each of the FIS experiment enclosures at a rateof 120 L d−1. The enclosures were sustained at 20 ◦C, a salinityof 33 psu, and a light intensity of 50 �E m−2 s−1 under a 16 h:8 hlight:dark cycle.

2.2. Methods used to monitor the biotic and abiotic factors

A sub-sample was collected from each microcosm at each spec-ified sub-sampling time (detailed in the SI Section 4). The watertemperature, pH, salinity, and dissolved oxygen (DO) concentra-tion in each sub-sample were measured using a portable meter(556 MPS; YSI, USA). A 100 mL aliquot of the sub-sample was fil-tered through a glass fiber filter (GF/F, Whatman) and stored inan acid-cleaned polyethylene bottle in a −80 ◦C deep freezer andthen the concentrations of inorganic nutrients in this sample weredetermined. The inorganic nutrients that were determined werethe dissolved inorganic nitrogen (DIN; NO2

− + NO3− + NH4

+), dis-solved inorganic phosphorus (DIP), and dissolved silica (DSi). Thenutrient concentrations in each sample were analyzed using anautomatic nutrient analyzer (Lachat Quickchem; Lachat Instru-ments, Milwaukee, USA). A 10 mL aliquot of each water samplewas filtered through a GF/F filter, and the dissolved organic carbon(DOC) concentration was determined using a high-temperaturecatalytic combustion instrument (TOC-VCPH; Shimadzu, Kyoto,Japan). These analytical methods were modified versions of themethods published by Grasshoff et al. [17].

To count bacterioplankton, a 30 mL aliquot was collected fromeach sub-sample in a 50 mL sterilized polyethylene bottle, fixedimmediately with glutaraldehyde (at a final concentration of 2%).The samples were stored in the dark at 4 ◦C until they were ana-lyzed. The fixed bacterioplankton cells were filtered through a blackpolycarbonate filter (GTBP 2500; Millipore, USA) and stained with a

1 �g mL−1 DAPI (4′,6-diamidino-2-phenylinodole) solution [18]. Atleast 600 stained bacterioplankton cells per sample were countedusing an epifluorescence microscope (Axioplan microscope; Zeiss,Germany) at a magnification of 1000×.

180 S.W. Jung et al. / Journal of Hazardous Materials 298 (2015) 178–187

Table 1The purposes of the experiments and the microcosm settings that were used in the study.

Experiment Purpose Inflow velocity ofwater

Numbers of enclosure Schematic diagram of microcosm setting used

PCS - Algicidal effects of HABs(C. polykrikoides and C.marina) and non-harmfulphytoplankton community- Responses of planktoniccommunities- Changes inenvironmental factors

None 6 enclosures(control: 3, treatment 3)

SIS - Algicidal effects of C.marina and non-harmfulphytoplankton community- Impact of Tetraselmis chuico-cultured with C. marina- Responses of planktoniccommunities- Changes inenvironmental factors

60 L d−1

(Slow inflow)4 enclosures(control: 2, treatment 2)

FIS - Algicidal activity of C.marina and non-harmfulphytoplankton community- Responses of planktoniccommunities- Changes inenvironmental factors

120 L d−1

(Fast inflow)2 enclosures(treatment 2)

The control groups are shown in blue color and the treatment groups in red color. PCS, perfectly closed system; SIS, slow inflow system (at a rate of 60 L d−1); FIS, fast inflowsystem (at a rate of 120 L d−1); CGCp, Cochlodinium polykrikoides control group; TGCp, C. polykrikoides treatment group; CGCm, Chattonella marina control group; TGCm, C.m up; TT m) an

aiasacfl

c2L

arina treatment group; CGPC, non-harmful phytoplankton community control grohe re-occurrence of Tetraselmis chui blooms when co-cultured with C. marina (TGC

To determine viability of C. marina and C. polykrikoides, a 30 mLliquot was collected from each sub-sample in a 50 mL steril-zed polyethylene bottle, and examined immediately (without anyddition of the fixative) using an inverted microscope (AxioOb-erver Z1; Zeiss, Germany) at a magnification of 100×–400× under

cold condition at temperature of 8–10 ◦C. HABs cells were notounted as dead if they did destruction of cell and no movement ofagella.

To count and identify the natural phytoplankton and ciliateells, a 200 mL aliquot was collected from each sub-sample in a50 mL sterilized polyethylene bottle, fixed immediately with acidugol’s fixative (at a final concentration of 2%). The fixed samples

GPC, non-harmful phytoplankton community treatment group.d blooms in non-harmful phytoplankton (TGPC) was determined in the SIS.

were concentrated via sedimentation. At least 500 phytoplank-ton cells and 100 ciliate cells per sample were examined usinga Sedgwick–Rafter counting chamber under a light microscope(Axioplan) at a magnification of 400×.

To count the heterotrophic nanoflagellates (HNFs), a 100 mLaliquot of each sub-sample was preserved with 25% buffered glu-taraldehyde (at a final concentration of 1%) and HNFs in the samplewere measured. The samples were stored in the dark at 4 ◦C prior to

analysis. At least 200 HNF cells were counted using an epifluores-cence microscope at a magnification of 400×–1000× after the cellshad been stained using the primuline staining method and excitedwith ultra violet light [19].

rdous

2

amwtspboTiatwa1mObua

2

uppwbcasS

3

3

a2Tttmgcitabc(tieafoptd

S.W. Jung et al. / Journal of Haza

.3. Elemental analysis of the white particles

White particles formed in all the treatment groups when the Ca-minoclay was added. We used a field emission scanning electronicroscope (FE-SEM; JSM-7600F; Jeol, Tokyo, Japan) to inspect thehite particles and used an EDS (X-Max Oxford, UK) attached to

he FE-SEM instrument to perform a qualitative elemental analy-is. A sub-sample of 50 mL was obtained from each microcosm. Thearticles in each water sample were collected on an Isopore mem-rane filter (pore size: 0.22 �m, GTTP04700; Millipore). The depthf particles collected on the membrane was approximately 1 mm.he samples were dried in a desiccator. Each dried sample was plat-

num (Pt) coated and then inspected using the FE-SEM instrumentt magnification of 8000×. To analyze the elements using the EDS,he takeoff angle was fixed at 40◦, and all of the visual and EDSork was performed at 15 keV with an 8 mm working distance

nd a prove current of 11 (gun emission current over the range00–105 �A). The analyses were performed in triplicate. Measure-ents of Pt were removed from the analysis software of EDS (Aztec,xford, UK) because of the Pt coating used to allow the samples toe visualized using FE-SEM. The total number of EDS input countssed to analyze a sample was approximately 300,000 over 15 min,nd the results were considered to be accurate to within ±1%.

.4. Statistical interpretation of the data

The results from different experimental groups were comparedsing the one-way analysis of variance method and then Scheffe’sost hoc test to assess the patterns in the decreasing HAB speciesopulation when the six different concentrations of Ca-aminoclayere present, and paired t-test was used to examine the difference

etween control and treatment. p values of less than 0.05 wereonsidered to indicate significant differences. Pearson’s correlationnalysis was used to examine the relationships between the mea-ured parameters. These statistical analyses were performed usingAS v.8.12 (SAS Institute Inc., Cary, NC, USA).

. Results

.1. Algicidal activities of Ca-aminoclay

The Ca-aminoclay completely destroyed the C. polykrikoidesnd C. marina cells when it was added at a concentration of50 mg L−1 in all the experiments (PCS, SIS, and FIS) (Fig. 1, SIable S1). The Ca-aminoclay did not disrupt the non-harmful phy-oplankton community in all the experiments (Fig. 1). However,he species compositions in the non-harmful phytoplankton com-

unity changed slightly differently in the control and treatmentroups during the experiments. Centric diatoms (mostly Chaeto-eros curvisetus) contributed more than 90% of the total communityn the CGPC over the experimental period, whereas, their con-ribution in the TGPC gradually decreased from 95.0% to 86.1%nd the pennate diatom (mostly Cylindrotheca closterium) contri-ution increased from 1.2% to 10.3% (SI Fig. S8). In the SIS, a T.hui that was co-cultured in the microcosms containing C. marinaTGCm) increased rapidly in abundance, from 0.8 × 104 cells mL−1

o 29.2 × 104 cells mL−1, when C. marina was destroyed by comingnto contact with the Ca-aminoclay (Fig. 1). Of particular inter-st in the SIS was a T. chui that was introduced in the seawaterdded to the TGCm and increased rapidly in abundance in the TGPCrom 0 to 30.6 × 104 cells mL−1, and the increase in the abundance

f this organism in the TGCm followed a similar trend (r = 0.99,

< 0.001). However, the green alga remained at a low cell density inhe CGCm and varied little throughout the experiment. The abun-ance of C. marina in the FIS TGCm experiment decreased rapidly,

Materials 298 (2015) 178–187 181

while the abundance of the non-harmful phytoplankton continu-ally increased throughout the experiment.

3.2. Changes in environmental factors

The water temperature and salinity in the control and treatmentgroups in the PCS and SIS experiments were little changed duringexperimental periods but the pH and DO concentrations in the PCSand SIS experiments were different (Fig. 2, SI Table S2, and Fig. S9).The pH in all of the PCS treatment groups increased immediatelyafter the Ca-aminoclay was added and then decreased gradually,although in the SIS and FIS experiments the pH decreased gradu-ally and then recovered to reach the initial values. The pH remainedconstant in all the control groups. The DO concentration decreasedrapidly when the Ca-aminoclay was added and then the conditionsremained anoxic in all of the PCS treatment groups. However, thechanges in DO concentrations differed in the SIS and FIS experi-ments to those in the PCS experiment. The DO concentration inthe SIS TGCm fell rapidly to almost zero after the experiment hadbeen running for 1.5 d, then remained constant for approximately4 d, and then increased until it reached the initial concentration.The DO concentration in the FIS experiment showed that a shorteranoxic shock occurred than in the SIS experiment. These changesmight have been affected by the volume and rate of the inflowingnatural seawater in these experiments. The variations in the DOconcentrations in the control groups were not notably different.

The changes in the nutrient and DOC concentrations thatoccurred during the different experiments were different (Fig. 2,SI Table S2 and Fig. S10). The DIN concentration in all of the PCStreatment groups increased rapidly for 1 d after the Ca-aminoclayhad been added, remained constant for 3 d, and then graduallydecreased. In particular, the DIN concentrations became higher inthe TGCp and TGCm experiments than in the TGPC experiment.The DIN concentrations changed less in the SIS and FIS experi-ments than in the PCS experiment. The DIP concentrations in thePCS TGCp and TGCm experiments continually increased once theCa-aminoclay had been added but decreased slightly in the TGPCexperiment. The DIP concentration decreased in each of the PCScontrol groups. In particular, DIP concentrations associated withblooms of C. polykrikoides and C. marina were detected close tothe start of the experiments. The DIP concentrations increased andthen gradually decreased in the SIS and FIS TGCm experiments,and the DIP concentrations were similar to the concentration inthe seawater being added to the microcosms by the end of theexperiments. The DSi concentrations followed similar patterns tothe DIN concentrations, increasing immediately in all of the exper-iments when the Ca-aminoclay was added and remaining constant(PCS), decreasing gradually (SIS), or decreasing rapidly (FIS). The DSiconcentrations remained constant in all of the control groups, andno particular differences were found between the different con-trol groups. The DOC concentrations also changed in similar waysto the DIN concentrations. In particular, the DOC concentrationsin all the treatment groups increased by a factor of approximately100 relative to all the control groups when the Ca-aminoclay wasadded. However, no particular differences were found between thechanges that occurred in the DOC concentrations in the controlgroups.

White particles formed in all the treatment groups when the Ca-aminoclay was added. The FE-SEM photographs indicated that theparticles had flat surfaces and that there were aggregations of smallparticles when viewed more closely (Fig. S12 and Table S3). The

EDS analysis results showed that the enrichment of some elements,such as C, O, Na, S, Cl, K, and Ca, was variable in the particles. Theelements C, O, and Cl made the highest contributions to the particlesby weight, at 76.04 wt%, 14.76 wt%, and 6.44 wt%, respectively.

182 S.W. Jung et al. / Journal of Hazardous Materials 298 (2015) 178–187

Fig. 1. (a) Changes in HAB species and non-harmful phytoplankton community abundances after inoculation of the Ca-aminoclay. The control groups are shown in blue colorand the treatment groups in red color. (b) Photographs of the microcosm experiments in which attempts were made to control HABs. The green colors in the SIS TGCm andTGPC photographs indicate that Tetraselmis chui blooms had occurred. (For interpretation of the references to colour in this figure legend, the reader is referred to the webversion of this article.)

S.W. Jung et al. / Journal of Hazardous Materials 298 (2015) 178–187 183

Fig. 2. Changes in the dissolved oxygen (DO) concentration, pH, dissolved inorganic nitrogen (DIN) concentration, dissolved inorganic phosphorus (DIP) concentration,dissolved silica (DSi) concentration, and dissolved organic carbon (DOC) concentration when the systems were treated with the Ca-aminoclay in microcosms. The controlgroups are shown in blue color and the treatment groups in red color. The error bars show the upper and lower 95% confidence intervals for the top and bottom points in eachb lues, ra cant d( erred

3

ttTmdwtodttPaae

s

ox plot. The closed circles and lines within the boxes are the mean and median vand Scheffe’s post hoc tests. Letters (A–D) and t-test (FIS experiment) indicate signifiFor interpretation of the references to colour in this figure legend, the reader is ref

.3. Response of the planktonic communities to Ca-aminoclay

The responses of the planktonic communities to the presence ofhe Ca-aminoclay, and the differences between the communities, inhe three experiments (PCS, SIS, and FIS) are shown in Fig. 3 and in SIable S2, and Fig. S11. The bacterioplankton cell abundances in theicrocosms varied in the opposite way to the HAB species abun-

ances. The bacterioplankton cell abundance rapidly increasedhen the Ca-aminoclay was added, reaching higher abundances

han were found in the control groups. In the PCS, the bacteri-plankton abundances in the TGCp and TGPC increased rapidlyuring the first 1–2 days of the experiment, then decreased, andhen increased again and remained constant for the remainder ofhe experiment. However, the bacterioplankton abundances in theCS TGCm continually increased in the absence of predators suchs HNFs and aloricate ciliates. The bacterioplankton abundances in

ll of the control groups did not vary significantly throughout thexperiments.

The bacterioplankton density increased and then the HNF den-ity increased gradually during the experiments. The time lag

espectively. Results were analysed by one-way ANOVA (PCS and SIS experiments)ifferences among experimental groups (p < 0.05 and p < 0.01). N.S is no significance.to the web version of this article.)

before the HNF abundances increased might have been caused bythe fluctuations in the total density of all of the bacterioplank-ton rather than by the impact of the Ca-aminoclay. The changesin the abundances of the ciliates might have been caused by fluc-tuations in the total bacterioplankton and HNF abundances. Inparticular, the abundances of predominant ciliates, Euplotes van-nus and Uronema sp., increased continually in the experiments inwhich the Ca-aminoclay was added (SI Fig. S13).

4. Discussion

4.1. Algicidal effect of Ca-aminoclay

When the Ca-aminoclay was added to the microcosms it killedthe HAB species completely and selectively. It was possible to deter-mine the algicidal selectivity of the Ca-aminoclay for HAB species,

and our results were consistent with those of Lee et al. [16], whofound that the Ca-aminoclay selectively killed HAB species includ-ing C. polykrikoides and C. marina. The Ca-aminoclay is a selectivealgicide when applied in natural water systems. Importantly, it is

184 S.W. Jung et al. / Journal of Hazardous

Fig. 3. Changes in the abundances of bacterioplankton, heterotrophic nanoflag-ellates, and ciliates when the systems were treated with the Ca-aminoclay inmicrocosms. The control groups are shown in blue color and the treatment groupsin red color. The error bars show the upper and lower 95% confidence intervals forthe top and bottom points in each box plot. The closed circles and the lines withinthe boxes are the mean and median values, respectively. Results were analysed byone-way ANOVA (all data in PCS and data of bacterioplankton in SIS experiments)and Scheffe’s post hoc tests. Letters (A–C) and t-test (data of bacterioplankton in FIS,of heterotrophic nanoflagellates and ciliates in SIS experiments) indicate significantdpt

ebiomicsw

scem

ifferences among experimental groups (p < 0.01). N.S is no significance. (For inter-retation of the references to colour in this figure legend, the reader is referred tohe web version of this article.)

asy to apply the Ca-aminoclay directly to natural water systemsecause it dissolves easily in seawater and also stable for a long time

n seawater [16]. We found that adding a relatively small amountf Ca-aminoclay was sufficient to control the HAB species; further-ore, the Ca-aminoclay does not cost much to produce [16]. It

s therefore, important to confirm the optimal Ca-aminoclay con-entration required to achieve algicidal activity in natural waterystems to ensure that the effective control of HABs is achievedithout applying excessive amounts of Ca-aminoclay.

The Ca-aminoclay immediately killed the HAB species in both

tanding and running water. This means that the Ca-aminoclayan be applied in any natural conditions because it rapidly andffectively causes lysis. Generally, the time taken for an algicidalaterial to mitigate a HAB is a very important factor. The biological

Materials 298 (2015) 178–187

controls of HABs have many advantages, such as offering species-specific and environmentally benign methods of control. However,these methods have certain weaknesses. For example, the algicidalbacterium Pseudomonas fluorescens HYK0210–SK09 responds onlyslowly and needs time to adapt to an environment before becomingactive enough to control HABs, and it can be attacked by predatorsduring the habilitation period [7,8,20]. However, a specific con-trol agent like the Ca-aminoclay will act only on the target speciesand can therefore be used to target and suppress only outbreaksof the target species. It is necessary to consider the safety of theentire ecosystem when developing chemical agents for controllingHABs.

We were able to evaluate the potential for the occurrence of abloom of another phytoplankton species after a HAB species hadbeen controlled using the Ca-aminoclay (Fig. 1). The Ca-aminoclaycompletely and selectively controlled blooms of C. marina in mixedcell cultures containing a T. chui (in the SIS TGCm experiment).However, the population of the T. chui increased over time and thenbloomed because of the hypertrophic nutrients released from theionized Ca-aminoclay and from the destroyed C. marina cells. More-over, the green alga was able to penetrate the cleaned water (inthe SIS TGPC experiment) and form a bloom. As mentioned abovethe introduction, controlling HABs using yellow clay can only offera limited ability to mitigate blooms. Large amounts of yellow claywere used in an attempt to control a massive C. polykrikoides bloomin the coastal waters around South Korea in 2013, and althoughthe C. polykrikoides blooms disappeared, the population of anotherdiatom, Guinardia delicatula, rapidly increased (unpublished data).We cannot offer a definitive explanation for the increase in theG. delicatula population. However, we assume that increase inthe diatom (G. delicatula behaved opportunistically) caused bythe release of nutrients when the yellow clay was ionized. TheCa-aminoclay is a focused and powerful agent for the selectivemitigation of specific HABs. It can, however, lead to blooms ofother phytoplankton occurring in natural aquatic systems. TheCa-aminoclay should, therefore, be used very carefully in naturalsystems, and, the possibility of another bloom must be considered.

4.2. Impact of using Ca-aminoclay on the environment

The DO concentrations in the microcosms decreased abruptlywhen the Ca-aminoclay was added and the subsequent develop-ment of anoxic conditions. There are in fact two possible reasons,namely the degradation of the biota and chemical changes. Thedegradation of the HAB organisms could lead to a decrease in theDO concentration because of increased bacterioplankton activityand a reduction in photosynthetic activity. Large blooms of phyto-plankton eventually settle to the bed in Chesapeake Bay (USA), andheterotrophic bacterioplankton then process much of this organicmaterial, which can result in the depletion of oxygen and evenanoxia [21]. Regarding possible chemical changes, CaCO3 particlesmay be formed in a system when CO2 is converted into HCO3

–

because of the abundance of surface amine groups in the Ca-aminoclay, and CO3

2− is produced when Ca2+ ions are added orreleased [22]. Therefore, we concluded that the decreases in theDO concentrations were caused by both the crystallization of theoxygen ions and the biological degradation of the dead HABs.

The nutrient and DOC concentrations also varied between theexperiment types. The concentrations increased immediately afterthe Ca-aminoclay had been introduced at the initial stage of eachexperiment (PCS, SIS, and FIS). However, the concentrations in thePCS experiment then remained constant throughout the exper-

imental period, whereas the concentrations in the SIS and FISexperiments reverted to their original levels because these sys-tems were diluted by the natural seawater that was constantlybeing added. Increases in these concentrations could be caused

S.W. Jung et al. / Journal of Hazardous Materials 298 (2015) 178–187 185

Fig. 4. Conceptual schematic diagram of the ecological responses when the Ca-aminoclay is introduced under different ecosystem scenarios. (a) Responses in biotic- anda ming

b a is pe

boaitPcoaeac

biotic factors in the closed system and (b) in the open system. (c) Effect of re-blooiotic- and abiotic factors in clean water when water bloomed by another microalg

y (1) the ionization of the Ca-aminoclay or (2) the degradationf the dead HAB organism cells. The DIN and DSi are supplied bymmonium and silica ions released by the Ca-aminoclay dissolv-ng (SI Fig. S5). However, the DIP concentrations also increased inhe experiments even though the Ca-aminoclay did not contain

atoms. The increases in the DIP concentrations may have beenaused by the degradation of the HAB organism cells. The releasef DOC derived from the dead HAB organisms was estimated to be

major source of the increases in the DOC concentrations in thexperiments. It is common for the DOC concentration to increasefter a phytoplankton bloom has ended [23,24]. Thus, Ca-aminoclayould cause secondary pollution and environmental damage.

by another phytoplankton after termination of HABs blooms, and (d) responses innetrated.

Moreover, the high nutrient and DOC concentrations produced canbe a fundamental cause of further blooms and can cause changesin the structures of microbial loops.

4.3. Impact of Ca-aminoclay on plankton organisms

The most obvious effect of adding the Ca-aminoclay in this studywas the rapid stimulation of bacterioplankton growth, which was

similar to the mesocosm results [25,26]. The rapid bacterioplank-ton growth was probably caused by the immediate availability of aplentiful energy source caused by the presence of the Ca-aminoclay.We thought that dissolved Ca-aminoclay may stimulate the

1 rdous

biobtTtect[

llnttaftpmcHsTpoptbaci

daPrbtptCbeuftc

4n

cecbdcabc

a

[

[

86 S.W. Jung et al. / Journal of Haza

acterioplankton growth being higher in the treatment groups thann the control groups. This might be explained by an indirect effectf the Ca-aminoclay, such as the proliferation of pollutants. If theacterioplankton abundance increased because of a direct effect ofhe Ca-aminoclay, the abundances would increase similarly in theGPC, TGCp, and TGCm experiments, but in fact the bacterioplank-on abundance was lower in the TGPC experiment than in the otherxperiments. We conclude that changes in the bacterioplanktonommunities that are induced by phytoplankton blooms are relatedo environmental variables and reflect changes in ecological niches27].

The initial increase in the bacterioplankton abundance was fol-owed by an increase in the abundance of HNFs and ciliates, with aag period of approximately 1–3 d, so the HNFs and ciliates mightot have been directly affected by the Ca-aminoclay but rather byhe availability of bacterioplankton as a food source. We concludehat bacterioplankton could grow rapidly soon after the Ca-minoclay was added because of the abundance of energy sourcesrom the dead HAB organisms, and that the subsequent progress ofhe microbial cycle could result in an increase in the abundance ofredators that feed on bacterioplankton. We attempted to deter-ine why the incidence of HNFs and ciliates (micro-predators)

ould increase in eutrophic conditions. One explanation is thatNFs can use a wide range of food items, including pico- and femto-

ized organisms, colloid molecules, and particles of detritus [28,29].herefore, any type of small particle that can spontaneously formarticulate structures can be consumed by HNFs (which are omniv-rous), allowing them to circumvent bacteria and directly accessarts of the pool of DOC material [29,30]. Our results indicate thathe microbial cycle was temporarily enhanced in our microcosmsecause of the availability of energy from bacterioplankton to HNFsnd ciliates, particularly in the Ca-aminoclay treatment groups. Weonclude that bacterioplankton and HNFs can be tolerant to thempacts of the Ca-aminoclay.

In terms of the succession of non-harmful phytoplankton (mostiatoms), the predominant species changed slightly when the Ca-minoclay was added, from centric diatoms to pennate diatoms.hytoplankton from different systematic taxa can show differentesponses to and tolerances of stress factors. Pennate diatoms haveeen found to be more tolerant of a number of various stress fac-ors in mesocosm studies than centric diatoms. In particular, theennate C. closterium was found to be more abundant than cen-ric diatoms in mesocosms affected by an oil spill [26,31] and. closterium also became very abundant when C. polykrikoideslooms were destroyed in mesocosm studies [25]. Underwoodt al. [32] found that C. closterium could proliferate under nutrient-nbalanced conditions. Therefore, variations in tolerance to stress

actors (nutrients, water temperature, and salinity etc.) among phy-oplankton species will affect the succession of the phytoplanktonommunity and the ecosystem produced.

.4. Is the Ca-aminoclay a good algicidal material for treatingatural HABs, and it safe to use in natural ecosystems?

This study was designed to test the Ca-aminoclay both in alosed system (PCS) and in open systems (SIS and FIS) as well asffect of re-blooming by another phytoplankton. The Ca-aminoclayompletely and selectively killed the HAB species in closed systems,ut the presence of the dead material caused the environment toeteriorate, which could cause damage to the biota and ecologicalhanges in the microbial cycle. The Ca-aminoclay also immedi-tely controlled the HAB species and caused environmental and

iological disturbances in the open systems, but the environmentalonditions recovered and returned to their initial conditions (Fig. 4).

Taking these results into account, we found that the Ca-minoclay may be not a suitable chemical treatment for use in

[

[

Materials 298 (2015) 178–187

natural environments. An essential goal of this study was to demon-strate that HABs could be controlled safely, with marine organisms,and the environment in general being relatively unaffected. Thisoutcome does indeed occur in the ocean and is a significantcontrol in marine ecosystems. Identifying the biochemical basesfor ecological phenomena will help constrain the phenomenonwithin an ecosystem context. Performing such studies for newmaterials is important because there is considerable interest inunderstanding the mode of action of such materials in treatingtarget organisms in natural waters. In this study, we estimatedthe significance of trophic interactions to the population dynamicsand to environmental effects using various scenarios. The knowl-edge gained from the study will be important when incorporatedinto our understanding of the interactions between microorgan-isms in the microbial cycle. In addition, the further work is requiredin the development of more promising techniques in conjunctionwith algicidal agents that can overcome these obstacles. Also, theresearch in the future is needed to test chronic toxicology for longterm periods and to elucidate the mechanism of the selective algi-cidal action in the ecosystem.

Acknowledgments

This research was supported by the Public Welfare & SafetyResearch Program through the National Research Foundation ofKorea (NRF) funded by the Ministry of Science, ICT & FuturePlanning of Korea (Development of autonomous microalgal iden-tification system and feasibility analyses of developed HAB controltechnologies: PN66020) and supported by the Research Program ofKorea Institute of Ocean Science & Technology (A study on the newatlas of marine resources in the adjacent seas to Korea: PE9931A).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.jhazmat.2015.05.012

References

[1] D.M. Anderson, Turing back the harmful red tide, Nature 388 (1997) 513–514.[2] A. Chambouvet, P. Morin, D. Marie, L. Guillou, Control of toxic marine

dinoflagellate blooms by serial parasitic killers, Science 322 (2008)1254–1257.

[3] H.G. Kim, Mitigation and controls of HABs, in: E. Granéli, J.T. Turner (Eds.),Ecology of Harmful Algae, Ecological studies, Springer, Berlin, 2006, pp.327–338.

[4] G.M. Hallegraeff, A review of harmful algal blooms and their apparent globalincrease, Phycologia 32 (1993) 79–99.

[5] T.J. Smayda, Novel and nuisance phytoplankton blooms in the sea: evidencefor a global epidemic, in: E. Granéli, B. Sundström, L. Edler, D.M. Anderson(Eds.), Toxic Marine Phytoplankton, Elsevier, New York, 1990, pp. 29–40.

[6] G.A. Rounsefell, J.E. Evans, Large Scale Experimental Test of Copper Sulfate as aControl for the Florida Red Tide, US Fish and Wildlife Service, Arlington, 1958.

[7] S.W. Jung, Y.-H. Kang, T. Katano, B.-H. Kim, S.-Y. Cho, J.H. Lee, Y.-O. Kim, M.-S.Han, Testing additions of Pseudomonas fluorescens HYK0210-SK09 to mitigateblooms of the diatom Stephanodiscus hantzschii in small- and large-scalemesocosms, J. Appl. Phycol. 22 (2010) 409–419.

[8] S.W. Jung, Y.-H. Kang, S.H. Baek, D. Lim, M.-S. Han, Biological control ofStephanodiscus hantzschii (Bacillariophyceae) blooms in a field mesocosm bythe immobilized algicidal bacterium Pseudomonas fluorescens HYK0210-SK09,J. Appl. Phycol. 25 (2013) 41–50.

[9] G.M. Rand, S.R. Petrocelli, Fundamentals of Aquatic Toxicology: Methods andApplications, Hemisphere, Washington, 1985.

10] R.L. Graney, J.H. Kennedy, J.H. Rodgers, Aquatic Mesocosm Studies inEcological Risk Assessment, CRC Press, Baca Raton, 1993.

11] T. Avnimelech, B.W. Troeger, L.W. Reed, Mutual flocculation of algae and clay:evidence and implications, Science 216 (1982) 63–65.

12] A. Shirota, Red tide problem and countermeasures (2), Int. J. Aquat. FishTechnol. 1 (1989) 195–293.

13] M.-C. Archambault, V.M. Bricelj, J. Grant, D.M. Anderson, Effects of suspendedand sedimented clays on juvenile hard clams Mercenaria mercenaria, withinthe context of harmful algal bloom mitigation, Mar. Biol. 144 (2004) 553–565.

rdous

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[oil on marine microbial communities in short term outdoor microcosms, J.Microbiol. 48 (2010) 594–600.

[32] G.J.C. Underwood, M. Boulcott, C.A. Raines, K. Waldron, Environmental effects

S.W. Jung et al. / Journal of Haza

14] A.J. Patil, M. Li, E. Dujardin, S. Mann, Novel bioorganic nanostructures basedon mesolamellar interaction or single-molecule wrapping of DNA usingorganoclay building blocks, Nano Lett. 7 (2007) 2660–2665.

15] Y.-C. Lee, W.K. Park, J.W. Yang, Removal of anionic metals byaminoorganoclay for water treatment, J. Hazard. Mater. 190 (2011) 652–658.

16] Y.-C. Lee, E.S. Jin, S.W. Jung, Y.-M. Kim, K.S. Chang, J.-W. Yang, S.-W. Kim, Y.-O.Kim, H.-J. Shin, Utilizing the algicidal activity of aminoclay as a practicaltreatment for toxic red tides, Sci. Rep. 3 (2013) 1–8.

17] K. Grasshoff, K. Kremling, M. Enrhordt, Methods of Seawater Analysis, seconded., Wiley-VCH, New York, 1985.

18] K.G. Porter, Y.S. Feig, The use of DAPI for identification and counting aquaticmicroflora, Limnol. Oceanogr. 25 (1980) 943–948.

19] D.A. Caron, Technique for enumeration of heterotrophic and phototrophicnanoplankton using epifluorescence microscopy, and comparison with otherprocedures, Appl. Environ. Microbiol. 46 (1983) 491–498.

20] S.W. Jung, B.H. Kim, T. Katano, D.S. Kong, M.-S. Han, Pseudomonas fluorescensHYK0210-SK09 offers species-specific biological control of winter algalblooms caused by freshwater diatom Stephanodiscus hantzschii, J. Appl.Microbiol. 105 (2008) 186–195.

21] F.K. Shiah, H.W. Ducklow, Temperature regulation of heterotrophicbacterioplankton abundance, production, and specific growth rate inChesapeake Bay, Limnol. Oceanogr. 39 (1994) 1243–1258.

22] D. Song, Y.C. Lee, S.B. Park, J.I. Han, Gaseous carbon dioxide convertion andcalcium carbonate preparation by magnesium phyllosilicate, RSC Adv. 4(2014) 4037–4040.

23] G. Bratbak, A. Jacobsen, M. Heldal, Viral lysis of Phaeocystis pouchetii and

bacterial secondary production, Aquat. Microb. Ecol. 16 (1998) 11–16.

24] A. Larsen, G.A.F. Flaten, R.-A. Sandaa, T. Castberg, R. Thyrhaug, S.R. Erga, S.Jacquet, G. Bratbak, Spring phytoplankton bloom dynamics in Norwegiancoastal waters: microbial community succession and diversity, Limnol.Oceanogr. 49 (2004) 180–190.

Materials 298 (2015) 178–187 187

25] S.H. Baek, M. Son, S.W. Jung, D.H. Na, H. Cho, M. Yamaguchi, S.-W. Kim, Y.-O.Kim, Enhanced species-specific chemical control of harmful and non-harmfulalgal bloom species by the thiazolidinedione derivative TD49, J. Appl. Phycol.26 (2014) 311–321.

26] S.W. Jung, O.Y. Kwon, C.K. Joo, J.H. Kang, M. Kim, W.J. Shim, Y.-O. Kim,Stronger impact of dispersant plus crude oil on natural plankton assemblagesin short-term marine mesocosms, J. Hazard. Mater. 217–218 (2012) 338–349.

27] H. Teeling, B.M. Fuchs, D. Becher, C. Klockow, A. Gardebrecht, C.M. Bennke, M.Kassabgy, S. Huang, A.J. Mann, J. Waldmann, M. Weber, A. Klindworth, A. Otto,J. Lange, J. Bernhardt, C. Reinsch, M. Hecker, J. Peplies, F.D. Bockelmann, U.Callies, G. Gerdts, A. Wichels, K.H. Wiltshire, F.O. Glöckner, T. Schweder, R.Amann, Substrate-controlled succession of marine bacterioplanktonpopulations induced by a phytoplankton bloom, Science 336 (2012) 608–611.

28] J.M. González, C.A. Suttle, Grazing by marine nanoflagellates on viruses andvirus-sized particles-ingestion and digestion, Mar. Ecol. Prog. Ser. 94 (1993)1–10.

29] E.B. Sherr, Direct use of high molecular weight polysaccharide byheterotrophic flagellates, Nature 335 (1988) 348–351.

30] H. Arndt, D. Dietrich, B. Auer, E.J. Cleven, T. Gräfenhan, M. Weitere, A.P.Mylnikov, Functional diversity of heterotrophic flagellates in aquaticecosystems, in: B.S.C. Leadbeater, J.C. Green (Eds.), The Flagellates: Unity,Diversity and Evolution, Taylor and Francis, London, 2003, pp. 240–268.

31] S.W. Jung, J.S. Park, O.Y. Kwon, J.H. Kang, W.J. Shim, Y.-O. Kim, Effects of crude

on exopolymer production by marine benthic diatoms: dynamics, changes incomposition, and pathways of production, J. Phycol. 40 (2004) 293–304.