Could the presence of larger fractions of non-cyanobacterial species be used as a predictor of...

$: Could the presence of larger fractions of non-cyanobacterial species be used as a predictor of microcystin production under variable nutrient regimes?$
Could the presence of larger fractions of non-cyanobacterialspecies be used as a predictor of microcystin productionunder variable nutrient regimes?

Som Cit Sinang & Elke S. Reichwaldt &Anas Ghadouani

Received: 15 January 2015 /Accepted: 16 June 2015# Springer International Publishing Switzerland 2015

Abstract The occurrence of cyanobacteria andmicrocystin is highly dynamic in natural environmentsand poses one of the biggest challenges to water resourcemanagement. While a number of drivers are known to beresponsible for the occurrence of cyanobacterial blooms,the drivers of microcystin production are not adequatelyknown. This study aims to quantify the effects of thechanges in the structures of phytoplankton andcyanobacterial communities on the dynamics ofmicrocystin production under highly variable nutrient con-centration. In our study, nutrient variability could explain64 % of the variability in microcystin production. Whenchanges in the fractions of non-cyanobacteria versuscyanobacteria genera were additionally included, 80 %of the variability in microcystin production could be ex-plained; under high nutrient concentrations, non-cyanobacterial phytoplankton groups were dominant overcyanobacteria and cyanobacteria produced more toxins. Incontrast, changes in the cyanobacterial community struc-tures could only explain a further 4 % of the dynamics ofmicrocystin production. As such, the dominance of non-cyanobacterial groups appears to be a useful factor to

explain microcystin occurrence in addition to traditionallyused factors such as absolute cyanobacterial cell numbers,especially when the nutrient regime is taken into account.This information could help to further refine the riskassessment frameworks which are currently used to man-age the risk posed by cyanobacterial blooms.

Keywords Phytoplankton . Cyanobacteria .

Microcystin production . Allelopathic interaction .

Nutrients . Public health

Introduction

An adequate understanding on the dynamics ofmicrocystin production will lead to a better managementof cyanobacterial blooms and reduce their impact onwater resources. While the occurrence of algal bloomsis considered to be a significant threat to water safetyand security around the world, the presence of high levelof cyanotoxins, such as microcystins, significantly in-crease the risk to the environment and the public. Assuch, it is critical to understand the key drivers behindthe toxin production during a bloom situation.

The dynamics of microcystin production in a bloomdepends on the toxin production on a cellular level,caused by changes in environmental factors such asnutrients and light (Chorus and Bartram 1999; Van deWaal et al. 2009; Zurawell et al. 2005), and has beenshown to be associated with the changes in the plank-tonic communities’ structure, in particular phytoplank-ton populations, including cyanobacterial as well as

Environ Monit Assess (2015) 187:476 DOI 10.1007/s10661-015-4695-z

S. C. Sinang : E. S. Reichwaldt :A. Ghadouani (*)Aquatic Ecology and Ecosystem Studies, School of Civil,Environmental, and Mining Engineering, The University ofWestern Australia, 35 Stirling Highway, M015, Crawley, WA6009, Australiae-mail: [email protected]

Present Address:S. C. SinangFaculty of Science and Mathematics, Sultan Idris EducationUniversity, 35900 Tanjong Malim, Perak, Malaysia

http://crossmark.crossref.org/dialog/?doi=10.1007/s10661-015-4695-z&domain=pdf

non-cyanobacterial species (Holland and Kinnear 2013;Joung et al. 2011; Pflugmacher 2002; Yang et al. 2014),and with changes in the ratio of toxic to non-toxicgenotypes (Janse et al. 2005; Kardinaal et al. 2007). Todate, many causation factors leading to changes in thecommunity structure of phytoplankton populations andthe development of cyanobacterial blooms have beenproposed and debated (Dadheech et al. 2014; Macíaset al. 2008; O’Neil et al. 2012; Paerl and Otten 2013).

Allelopathy has been suggested as a key mechanismin determining the community structure and the dynam-ics of phytoplankton populations (Holland and Kinnear2013; Macías et al. 2008). The theory of allelopathysuggests that the ability of certain phytoplankton groupsto dominate the community may be enhanced by theproduction of toxic or non-toxic compounds (Camacho2008; Graneli and Hansen 2006; Li and Li 2011; O’Neilet al. 2012). This theory has been used in an attempt toexplain the ecological role of microcystin production incyanobacteria; however, this has not been extensivelysupported by experimental data (Babica et al. 2006;Berry et al. 2008; Jang et al. 2006). As such,cyanobacteria species could increase their competitiveadvantage through the production of microcystin(Huisman and Hulot 2005), especially under certainconditions of nutrient limitation (Graneli et al. 2008).Phytoplankton exposure to microcystin has been sug-gested to be associated with growth inhibition and de-terioration of photosynthesis and enzymatic activities insome non-cyanobacteria species (Babica et al. 2006;O’Neil et al. 2012). Consequently, it can be hypothe-sized that the dynamics of microcystin production isrelated to the inter-species interaction betweencyanobacteria and other phytoplankton species, interac-tion within cyanobacterial communities, as well as thedirect effects of environmental factors on phytoplank-ton. While the concept of allelopathy is hard to prove insitu, as the effect of the chemical on the receivingorganism is alleviated by the effects of other environ-mental factors, an increase in the production of toxins bycyanobacteria in the presence of a larger fraction of non-cyanobacterial genera might be an indirect indicator forallelopathy. Many studies have been focusing on theeffect of nutrients on microcystin production (Chorusand Bartram 1999; Orihel et al. 2012; Zurawell et al.2005); while a limited number of studies have alsoinvestigated the combined effects of phytoplanktoncommunity structure and cyanobacterial species com-position on microcystin production (Engström-Öst et al.

2011; Kardinaal et al. 2007; Millie et al. 2009), none hasstudied this in relation to changes in nutrient levels. Thisinformation is essential and would enable managers toassess the potential toxicity of cyanobacterial bloomsfrom the structure of the phytoplankton community andcyanobacterial composition under different environ-mental conditions. The present study aims to quantifythe effects of the changes in the structure of phytoplank-ton communities, including cyanobacterial species com-position, on the dynamics of microcystin productionunder highly variable nutrient regimes. It is anticipatedthat nutrient dynamics, specifically phosphorus, nitro-gen, and iron, alone are not good predictors ofcyanotoxin production and that key changes in the phy-toplankton communities, especially a dominance ofnon-toxin producing classes, such as cholorophytes ordiatoms, may be responsible for triggering anallopathetic response leading to a higher toxinproduction.

Materials and methods

Study area

This study was carried out in Jackadder Lake (31° 54′30″ S, 115° 47′ 36″ E), Bibra Lake (32° 5′ 25″ S, 115°49′ 16″ E), and Yangebup Lake (32° 6′ 56″ S, 115° 49′33″ E) located on the Swan Coastal Plain, WesternAustralia. These lakes are shallow with mean depth of2.1, 1.1, and 2.5 m in Jackadder Lake, Bibra Lake, andYangebup Lake, respectively. Jackadder Lake issurrounded by 6.6 ha of parkland (Arnold 1990;Woodward 2008), while Bibra Lake is surrounded byurban areas and a golf course. Yangebup Lake issurrounded by residential, agriculture, and industrialareas. Jackadder Lake and Yangebup Lake are perma-nent lakes, while Bibra Lake is subject to seasonaldrying. Prolonged stable thermal stratification is usuallyprevented in these lakes during summer due to contin-uous or intermittent wind mixing that creates a homo-geneous environment throughout the water column(Arnold and Oldham 1997; Davis et al. 1993). Theselakes were selected due to levels of cyanobacterial bio-mass and microcystin concentration reported in earlierextensive sampling (Reichwaldt et al. 2013; Sinanget al. 2013). These lakes represent systems with low,medium, and high cyanobacterial biomass andmicrocystin concentration. Mean cyanobacterial

476 Page 2 of 15 Environ Monit Assess (2015) 187:476

biomass in an earlier study was reported as 28 μg chl-aL−1 in Jackadder Lake, 108 μg chl-aL−1 in Bibra Lake,and 80 μg chl-aL−1 in Yangebup Lake. Mean cellularmicrocystin concentrations (mg g−1 cyanobacterial drymass) was 4.8 mg g−1 in Jackadder Lake, 35 mg g−1inBibra Lake, and 1.7 mg g−1 in Yangebup Lake (Sinanget al. 2013).

Sampling

The study lakes were sampled twice a month betweenJanuary and March 2010. Samples were collected fromthree points on each lake on every sampling occasion.Samples were taken from waist-deep water at the samepoint on every sampling occasion. Bibra Lake dried upin late February; therefore, no samples were taken fromthis lake in March.

Grab water samples were taken from 15 cm belowthe surface and stored immediately in a cool container.Parameters analyzed from these samples were totalphosphorus (TP), total dissolved phosphorus (TDP),total iron (TFe), total dissolved iron (TDFe), total nitro-gen (TN), total dissolved nitrogen (TDN), ammonium(NH4

+), cyanobacterial biomass, total phytoplanktonbiomass, intracellular and extracellular microcystin frac-tions, and cellular microcystin concentration. Samplesfor dissolved nutrient analyses were pre-filtered with a0.45-μm syringe filter (Acrodisc, HT Tuffryn) beforestoring at −20 °C. Water samples for phytoplanktonenumeration were preserved with Lugol’s iodine (1 %final concentration).

Nutrient analyses and phytoplankton biomass

TP and TDP concentrations were analyzed using theascorbic acid method, while TFe and TDFe concentra-tions were analyzed with the phenanthroline method,according to standard methods (APHA 1998). TN,TDN, and NH4

+ were analyzed at the South CoastNutrients Analysis Laboratory (Albany, WesternAustralia) with the standard colorimetric methods on asegmented flow auto-analyzer (Alpkem, Wilsonville,OR, USA). Cyanobacterial and total phytoplanktonchlorophyll-a were measured with a benchtop versionof a FluoroProbe (bbeMoldaenke, Germany), which iscapable of identifying up to four spectral algal groupsbased on their specific excitation spectrum of chloro-phyll fluorescence (Beutler et al. 2002, Ghadouani andSmith 2005). The biomass of each group is given in

equivalent μg chl-aL−1. Fluoroprobe measurements oftotal chlorophyll-a gave a good correlation against sam-ples extracted according to standard methods (APHA1998) (linear regression analysis: R2=0.94, N=32, P<0.05) (Sinang, S.C., unpublished data).

Water samples were filtered through pre-combustedand pre-weighed 47-mm GF/C filter papers to collectthe phytoplankton biomass and to separate intracellularfrom the dissolved microcystin fraction. Filter paperscontaining particulate organic matter and phytoplanktonbiomass were dried at 60 °C and re-weighed after 24 h toobtain the dry biomass (Chorus and Bartram 1999).These filter papers were then moistened with Milli-Qwater and stored at −20 °C until intracellularmicrocystin extraction. As we were interested in theactual toxic potential of cyanobacteria, cyanobacterialdry mass was calculated from the total dry mass byadjusting it to the percentage of cyanobacterial biomassmeasured with the FluoroProbe.

Cyanobacteria were identified and enumerated to thetaxonomic level of genera with 10–50-mL water sam-ples fixed in Lugol’s solution using an inverse micro-scope according to Utermöhl (1958). A minimum of200 cells or colonies of the most abundant cyanobacteriawere counted, and the size of 20 individual cells orcolonies of each occurring genera was measured byusing an inverse microscope. Cell densities were con-verted to biovolume by means of geometric figures thatbest approximated the shape of individual taxa(Hillebrand et al. 1999). The biovolume of eachcyanobacterial genera was then used to calculate thebiovolume fraction of the respective cyanobacterial gen-era in the total cyanobacterial community.

Microcystin extraction and quantification

Filters were freeze-thawed twice to break the cells priorto methanol extraction (Lawton et al. 1994). Each filterwas placed into a separate centrifuge tube and 5 mL75 % methanol-water (v/v) was added. Filters weresonicated on ice for 25 min, followed by gentle shakingfor another 25min. The extracts were then centrifuged at3750 rpm for 10min at room temperature. Extracts werecarefully transferred into conical flasks, and two moreextractions were done per filter. All three extracts (75 %methanol) were pooled and diluted withMilli-Q to 20%methanol. Intracellular microcystin extracts and the pre-filtered water containing dissolved microcystin weresubjected to solid-phase extraction (SPE) (Waters

Environ Monit Assess (2015) 187:476 Page 3 of 15 476

Oasis HLB) as a cleanup step and for concentrating themicrocystins. Sample loading speed to the SPE wasadjusted to <10 mLmin−1. After the samples had passedthrough, each cartridge was rinsed with 10 mL 10, 20,and 30 % methanol-water (v/v), before microcystin waseluted with 100 % methanol+0.1 % trifluoroacetic acid(TFA) into 6-mL glass vials and samples were evapo-rated to dryness under a mild nitrogen stream in a 40 °Cwater bath. Finally, samples were re-dissolved in 30 %acetonitrile and analyzed with HPLC-PDA systems(Waters, Alliance 2695).

HPLC analysis was carried out by using the Alliance2695 (Waters, Australia) with a PDA detector (1.2-nmresolution) and an Atlantis T3 3-μm column (4.6×150 mm i.d.). Mobile phases used were acetonitrile+0.05 % v/v TFA and Milli-Q water+0.05 % v/v TFA.Microcystin peak separation was achieved by using alinear gradient a flow rate of 0.6 mL min−1 (ISO 2005).Column temperature was maintained at 37.5±2.5 °C.The limit of detection per microcystin peak was 1.12 ng.Microcystin was identified by PDA spectra at 238 nm.Commercially available microcystin-LR standard(Sapphire Bioscience, Australia; purity ≥95 %) wasused to quantify microcystin concentrations in allsamples.

Statistical analyses

Redundancy analysis (RDA) (Legendre andLegendre 1998) was carried out in R language toquantify the effects of nutrients, and the structure ofphytoplankton and cyanobacterial community onmicrocystin production. The RDA analysis was car-ried out using the combined data from three lakes aswe emphasize on finding the general trend ratherthan individual lake pattern. Explanatory variablesused in the RDA analysis were nutrient concentra-tions in the water column, the fraction of differentphytoplankton groups of total phytoplankton bio-mass, and the fraction of different cyanobacterialgenera of total cyanobacterial community. Responsevariables used in the RDA analysis were cellularmicrocystin concentration and total microcystin con-centration in the water column. The canonical rela-tionship was computed with standardized explanatoryand response variables and tested for the level ofsignificant after 999 permutations. In all analyses,results were considered significant at P<0.05.

Results

Nutrient profiles of studied lakes

The nutrient profiles in Jackadder, Bibra, and YangebupLakes during the study period are shown in Fig. 1a–h.Nutrient concentrations varied on a temporal basis with-in lakes and spatially between lakes. Phosphorus con-centrations were higher in Bibra Lake than in Jackadderand Yangebup Lakes throughout the sampling period.Mean TP concentrations (Fig. 1a) ranged from 22 up to~92, 230 to >1000, and 28 to >150 μg L−1 in Jackadder,Bibra, and Yangebup Lakes, respectively. Meanwhile,mean TDP concentrations (Fig. 1b) ranged from 12 to24, 17 to 142, and 14 to 37 μg L−1 in Jackadder, Bibra,and Yangebup Lakes, respectively.

TN (Fig. 1c) and TDN (Fig. 1d) concentrations inBibra Lake were higher by up to one order of magnitudethan concentrations in Jackadder and Yangebup Lakes.In contrast, mean molar TN/TP in Bibra Lake was lowerthan the ratios in Jackadder and Yangebup Lakes(Fig. 1e). Mean TN/TP ranged from 40 to 132, 36 to84, and 65 to 255 in Jackadder, Bibra, and YangebupLakes, respectively. NH4

+ decreased over time inJackadder and Yangebup Lakes (Fig. 1f). Mean NH4

+

concentrations ranged from 43 to 170, 157 to 239, and40 to 143 μg L−1 in Jackadder, Bibra, and YangebupLakes, respectively. Mean TFe (Fig. 1g) and TDFe(Fig. 1h) concentrations were higher in Bibra Lake thanin the other two lakes during the beginning of the study.Mean TFe ranged from 77 to 247, 147 to 220, and 51 to110 μg L−1 in Jackadder, Bibra, and Yangebup Lakes,respectively. Mean TDFe ranged from 24 to 174, 61 to117, and 21 to 89 μg L−1 in Jackadder, Bibra, andYangebup Lakes, respectively.

Dynamics of phytoplankton community, cyanobacterialcomposition, and microcystin production

The phytoplankton composition in Jackadder, Bibra,and Yangebup Lakes changed during the study period(Fig. 2). Phytoplankton composition varied on a tempo-ral basis within lakes and spatially between lakes. InJackadder Lake, cyanobacteria were dominant (68 %)from January to early February. In late February, thephytoplankton community shifted to mainlychlorophytes and diatoms, and cyanobacteria onlyaccounted for 7 % of the community. Then, in earlyMarch, cyanobacteria started to increase and contributed


Fig. 1 Mean nutrient concentrations (a TP, b TDP, c TN, d TDN, e TN/TP, f NH4+, g TFe, h TDFe) in Jackadder, Bibra, and Yangebup

Lakes from January to March 2010


to 67% of the total phytoplankton biomass by the end ofMarch. In Bibra Lake, chlorophytes dominated withthe biomass up to 66 %, while cyanobacteria onlycontributed 27 % to the total phytoplankton biomassin January. Cyanobacteria contributed up to 49 % ofthe total phytoplankton biomass in early February,and the dominant phytoplankton changed tochlorophytes in late February. In Yangebup Lake,cyanobacteria were dominant in January and

cyanobacteria contributed up to 67 % to the totalphytoplankton biomass. In February, cyanobacterialfraction in the total phytoplankton biomass decreasedfrom 64 to 17 %, and chlorophytes took over themajor fraction of total phytoplankton biomass in lateFebruary. In early March, cyanobacteria contributed48 % to the total phytoplankton biomass inYangebup Lake and the fraction decreased to 32 %in late March.

Fig. 2 Phytoplanktoncommunity composition (% oftotal) in a Jackadder, b Bibra, andc Yangebup Lakes from Januaryto March 2010. Bibra Lake driedup in early March; therefore, nosamples were taken from this lakein March


At the cyanobacterial genus level, cyanobacterialcomposition was highly variable between lakes, and,within a lake, in time. The cyanobacterial communityin Jackadder Lake was mainly composed of three po-tential microcystin-producing genera, namelyMicrocystis spp., Planktothrix spp., and Anabaena spp.(Fig. 3a). There was a shift of dominance fromMicrocystis spp. to Planktothrix spp. within a 7-dayperiod in January, after which Microcystis spp. wasagain dominant (Fig. 3a). Planktothrix spp. biovolumereached 2.208×103 mm3 mL−1 in late January and de-creased continuously towards the end of the study peri-od (Fig. 4a). From February until the end of the study

period, Microcystis spp. was the major fraction in thecyanobacterial community with biovolumes rangingfrom 0.318×103 to 1.896×103 mm3 mL−1 (Fig. 4a).Anabaena spp. was present as a small fraction inJanuary and February and increased to 50 % of the totalcyanobacterial community in March (Fig. 3a). In BibraLake, Anabaenopsis spp., Anabaena spp., Planktothrixspp., Nodularia spp., and Microcystis spp. were thecyanobacteria detected, and their percentage biomassvaried over time (Fig. 3b). In January, there was achange in dominance from Anabaenopsis spp. toMicrocystis spp. within a 7-day period (Fig. 3b).During its dominance in early January, Anabaenopsis

Fig. 3 Cyanobacterialcommunity composition (% oftotal) in a Jackadder, b Bibra, andc Yangebup Lakes from Januaryto March 2010. Bibra Lake driedup in early March; therefore, nosamples were taken from this lakein March


spp. reached a biovolume of 4.884×103 mm3 mL−1, butit was no longer detected after the end of February(Fig. 4b). Simultaneously, with the decrease ofAnabaenopsis spp., the fraction of Microcystis spp. in-creased from early January and remained the dominantcyanobacterial genera throughout the study period(Fig. 3b). Microcystis spp. biovolumes ranged from1.865×103 to 111.34×103 mm3 mL−1 (Fig. 4b).Anabaena spp. also coexisted, but the total biovolumewas always less than 50 % of total cyanobacteria bio-mass (Fig. 3b). The highest Anabaena spp. biovolumewas detected at 80.2×103 mm3 mL−1 in early February(Fig. 4b). Planktothrix spp. and Nodularia spp. contrib-uted less than 5 % of the total cyanobacterial biomass inBibra Lake (Fig. 3b). In Yangebup Lake, Microcystis

spp. was the dominant genera throughout the studyperiod (Fig. 3c). Microcystis spp. biovolumes rangedfrom 2.766×103 to 9.839×103 mm3 mL−1 (Fig. 4c).Planktothrix spp. and Nodularia spp. contributed lessthan 20 and 1 % to the total cyanobacterial biomass,respectively (Fig. 3c).

Cellular microcystin concentration (mg MC-LRequiv

g−1 cyanobacterial dry mass) and total microcystin con-centration in the water were highly variable over time inall lakes (Fig. 5). Cellular microcystin concentrationranged in three orders of magnitude in Jackadder Lakeand two orders of magnitude in Yangebup and BibraLakes. In all lakes, cellular microcystin concentrationincreased from January to February and decreased inMarch. The highest mean cellular microcystin

Fig. 4 Biovolume concentrations(mm3 mL−1) of cyanobacterialgenera over time in a Jackadder, bBibra, and c Yangebup Lakesfrom January to March 2010.Bibra Lake dried up in earlyMarch; therefore, no sampleswere taken from this lake inMarch


concentration occurred in late February in all lakes. Thehighest mean cellular microcystin concentration was1.4 mg g−1 dry mass in Jackadder Lake, 0.6 mg g−1

dry mass in Bibra Lake, and 0.5 mg g−1 dry mass inYangebup Lake. Total microcystin concentration in the

water varied by an order of magnitude in all lakes. InBibra Lake, total microcystin concentration increasedsignificantly from January to February, and mean con-centrations ranged between 0.7 μg L−1 in January to6.9 μg L−1 in February. In Jackadder Lake, mean total

Fig. 5 Mean cellular microcystinconcentration (mg g−1

cyanobacterial dry mass) andmean total microcystinconcentration in the water(μg L−1) in Jackadder, Bibra, andYangebup Lakes from January toMarch 2010; error bars representstandard errors


microcystin concentration ranged between 0.2 and1.3 μg L−1, and the highest concentration occurred inearly March. In Yangebup Lake, mean total microcystinvaried between 0.4 and 2.1 μg L−1, and the highestmean concentration was also detected in early March.

Correlations between nutrients, phytoplanktonand cyanobacterial community structure,and microcystin production

RDA analysis showed that the dynamics of microcystinproduction in the studied lakes were significantly corre-lated to nutrient concentrations in the water column.Nutrients explained 64 % of variation in cellularmicrocystin concentration (Fig. 6). Cellular microcystinconcentration was closely correlated to iron concentra-tion, while total microcystin concentration in the waterwas closely correlated to phosphorus and nitrogen avail-ability in the systems. Inclusion of cyanobacterial com-position as an explanatory variable in addition to nutri-ents increased the percentage explained from 64 to 68%(P<0.05), suggesting that nutrients were more impor-tant as a factor influencing the dynamics of cellularmicrocystin production than the cyanobacterial compo-sition (data not shown). In contrast, including the

phytoplankton community as an explanatory variablein addition to nutrients, 80 % of variability inmicrocystin production could be explained (P<0.05)(Fig. 7). The combination of the phytoplankton commu-nity and nutrients as explanatory variables increased theability to predict the dynamics of microcystin produc-tion by 16 %, indicating the significance of both vari-ables on microcystin production. As shown in the RDAbiplot (Fig. 7), highmicrocystin production correlated tohigh nutrient concentrations in the presence of a highfraction of chlorophytes and diatoms and a low fractionof cyanobacteria in the phytoplankton community. Inaddition, higher cellular and total microcystin concen-trations were associated with lower fractions ofcyanobacteria present in the systems (Fig. 8). Therewas a significant but very weak correlation betweentotal cyanobacterial biomass and total microcystin con-centration in the water (R2=0.175; P<0.05; y=0.99+0.05*x).

Discussion

Changes in water levels during dry months can havestrong effects on nutrient dynamics, which subsequently

Fig. 6 RDA biplot of cellular microcystin (cMC) and totalmicrocystin (tMC) concentration with nutrients; solid lines = en-vironmental variables; short dashed lines = response variables.Canonical axes 1 and 2 represent a linear combination of theenvironmental variables

Fig. 7 RDA biplot of cellular microcystin (cMC) and totalmicrocystin (tMC) with nutrients and phytoplankton community(Chloro = chlorophytes; Crypto = cryptophytes; Diatom; andCyano = cyanobacteria); solid lines = environmental variables;short dashed lines = response variables. Canonical axes 1 and 2represent a linear combination of the environmental variables


could affect the structure of phytoplankton communities inlakes (Coops et al. 2003; Jeppesen et al. 2015). As such,this study has taken into account both nutrient and phyto-plankton dynamics to understand the dynamics ofmicrocystin production. The results of our study suggestedthat high microcystin production in the studied lakes wassignificantly correlated to high nutrient concentration andlow TN/TP. Our results agree with previous studies, whichhave reported that increased phosphorus, nitrogen, andiron concentrations could enhance microcystin production(Albay et al. 2005; Jiang et al. 2008; Scott et al. 2014), andthat higher microcystin concentrations can be found atlower TN/TP ratios (Orihel et al. 2012). The observedclose positive relationships between microcystin produc-tion with phosphorus and nitrogen are potentially due tothe high cyanobacterial cellular requirement of phosphorusfor microcystin biosynthesis and biomass production(Vezie et al. 2002).

In addition, a positive correlation between iron andmicrocystin production may be related to an earlier sug-gestion that microcystin production involves enzyme syn-thetase, in which the enzyme’s activity is controlled by theamount of iron present (Utkilen and Gjolme 1995). Inaddition to the possible direct effect of nutrients on thedynamic of microcystin production, previous studies havealso suggested that nutrients might indirectly regulatemicrocystin production through their effects oncyanobacterial abundance and its distribution (Dai et al.2008; Millie et al. 2009; Wu et al. 2014). Unlikechlorophytes and diatoms, a high cyanobacterial fractionwas pointing in the opposite direction to high nutrientavailability (Fig. 7). This result indicates that high nutrientconcentrations in our systems did not favor the dominance

of cyanobacterial genera. Instead, high phosphorus, nitro-gen, and iron concentrations in thewater columnmay havetriggered the dominance of chlorophytes and diatoms, dueto their ability to grow faster than cyanobacteria (Huismanet al. 2005; Jensen et al. 1994; Reynolds et al. 2006).

Microcystin production is complex and various en-vironmental factors including nutrients have beenshown to have direct effects on the dynamics ofmicrocystin production (Rinta-Kanto et al. 2009). Ourstudy suggests that the dominance of non-cyanobacterial groups is an additional factor to be con-sidered and that changes in the structure of the phyto-plankton community may have a large effect onmicrocystin production. Changes in the structure of thephytoplankton community explained up to 16 % of thevariability in microcystin production. Moreover,cyanobacteria tend to increase microcystin productionin the presence of a high fraction of other phytoplanktongroups. This result supports the theory of allelopathicinteractions in cyanobacteria through the higher produc-tion of toxin microcystin with an increasing number ofcompetitors (Holland and Kinnear 2013; Huisman andHulot 2005; Rzymski et al. 2014). Earlier studies onallelopathy suggest that cyanobacteria may be capableof reacting to the presence of competitors by increasingmicrocystin production stimulated by the presence ofextracellular products released by competitors such aschlorophytes (Bittencourt-Oliveira et al. 2014; Kearnsand Hunter 2000; Pinheiro et al. 2013). In addition to theeffects of changes in the structure of phytoplanktoncommunity, the dynamics of microcystin productioncould also be related to the succession of toxic andnon-toxic cyanobacterial genotypes (Janse et al. 2005;

Fig. 8 Correlation betweencyanobacterial fraction andcellular microcystin concentration(mg g−1 cyanobacterial dry mass)(solid line; R2=0.542; P<0.05;y=2.85−1.78*x) or totalmicrocystin concentration(μg L−1) (dashed line; R2=0.105,P<0.05; y=2.96−3.08*x)


Kardinaal et al. 2007; O’Neil et al. 2012), which was notanalyzed in this study.

Based on the available cyanobacterial genera com-position data, our results suggest that the changes inthe structure of the cyanobacterial community didnot affect the dynamics of microcystin productionto a large extent. This result indicates that the suc-cession of Planktothrix spp., Anabaena spp.,Nodulariasp., Anabaenopsis spp., and Microcystisspp. in our lakes may not be related to changes inmicrocystin production in response to cyanobacterialcompetitors, and, as such would not support thetheory of allelopathic interaction between differentcyanobacteria genera (Paerl and Otten 2013;Švercel 2013). Rather, the succession of differentcyanobacteria genera may be related to differentcyanobacterial genera benefitting from different con-centrations of phosphorus, nitrogen, and iron species(Chorus and Bartram 1999; Huisman and Hulot2005). Drastic changes to the structure ofcyanobacterial communities are possible even withonly slight changes in the cyanobacterial total bio-mass (Soares et al. 2009). It was suggested that thiscould be due to the short doubling times ofcyanobacteria in addition to different cyanobacterialgenera occupying different niches, which developdue to temporal changes in the water’s physicochem-ical properties (Reynolds 1998).

As an example, Microcystis spp. are known to becapable to store high amounts of phosphorus and to usephosphorus very efficiently (Baldia et al. 2007), whilePlanktothrix spp. has a lower affinity toward phospho-rus (Istvánovics et al. 2002). In addition, Nagai et al.(2007) has suggested thatPlanktothrix agardhii requiresan iron concentration 30 times higher compared toMicrocystis aeruginosa to achieve the same growth rate.This indicates that cyanobacteria may react differentlyto environmental conditions created by other genera,which could lead to drastic changes in the structure ofthe cyanobacterial community over short periods of time(Beversdorf 2013). As an example, the higher iron re-quirement of Planktothrix spp. might reduce the ironavailability in the water column, and it would thusbecome favorable for Microcystis spp. in the system.Therefore, the succession of different cyanobacteriagenera is a driven by a complex interaction of environ-mental conditions, including the availability of nutri-ents, and allelopathic interactions through cyanotoxinsor other metabolites (Švercel 2013).

Conclusions

Our results indicate that the changes in the structure ofthe phytoplankton community and changes in nutrientconcentrations are important factors that determine thedynamics of microcystin production. Under high nutri-ent concentration, non-cyanobacterial phytoplanktongroups were dominant over cyanobacteria, andcyanobacteria produced more microcystin. Our resultsmight therefore support the theory of allelopathic inter-actions between cyanobacteria and non-cyanobacterialcompetitors.We conclude that eutrophication could leadto the occurrence of more toxic cyanobacterial bloomsdirectly through increased nutrient concentrations, andindirectly through changes in the phytoplankton com-munity leading to increased toxin production within thecyanobacterial cells.

Our resul ts also suggest that changes incyanobacterial community composition may not be ofmajor importance for the dynamics of microcystin pro-duction. In relation to the microcystin risk assessmentduring toxic cyanobacterial blooms, our results there-fore suggest to consider the fraction of totalcyanobacteria in addition to the absolute cyanobacterialbiomass, which is currently the basis of existing WHOalert levels for microcystin risk assessment (Churro et al.2012). Our conclusion is drawn based on the quantifi-cation of total biomass for each cyanobacterial generawithout differentiating between microcystin-producingstrains and non-producing strains. The quantification oftoxic and non-toxic cyanobacteria strains would haveallowed for a deeper understanding of the microcystindynamics and the dominance within the cyanobacteriacommunity and we therefore suggest to include molec-ular characterization of cyanobacterial communities infuture studies to get a clearer picture of the role of straindynamics on microcystin production, which would helpto develop a more assertive microcystin risk assessment.

Additionally, a direct relationship between the struc-ture of the phytoplankton/cyanobacteria community andmicrocystin production is likely to be established. Sucha relationship illustrates the possibility of using thecyanobacterial fraction in the phytoplankton communityas an ecological indicator to infer the level ofmicrocystin production or risk of microcystin contami-nation in shallow water bodies. The relationship ob-served in this study was in agreement with the theoryof allelopathic interaction between cyanobacteria andnon-cyanobacteria. As direct microcystin quantification


is not regularly conducted in water quality monitoring,the observation of high nutrient concentrations in thepresence of potentially toxic cyanobacterial species,together with the dominance of non-cyanobacterial phy-toplankton groups over cyanobacteria could be used asan indication that toxin production per cell is high andshould prompt direct microcystin quantification. Evenso, we propose further testing of this relationship with alarger number of lakes, including verification for singlelakes, to prove its application for lake management.

Acknowledgments This project was funded by the AustralianResearch Council’s Linkage Project funding scheme (LP0776571)and theWater Corporation of Western Australia. We wish to thankthe City of Cockburn and the City of Stirling for permission tosample the lakes, Professor Pierre Legendre, Laura Firth, andKevin Murray for their valuable statistical advice, and LiahCoggins for her help in the editing of the manuscript. During thestudy, Sinang, S.C was supported by a scholarship from UniversitiPendidikan Sultan Idris (UPSI) and the Malaysian Government.

References

Albay, M., Matthiensen, A., & Codd, G. A. (2005). Occurrence oftoxic blue-green algae in the Kucukcekmece Lagoon(Istanbul, Turkey). Environmental Toxicology, 20, 277–284.

APHA (1998). Standard methods for the examination of waterand wastewater 20th edn.

Arnold, J. (1990). Perth Wetlands Resource Book. Perth:Environmental Protection Authority.

Arnold, T. N., & Oldham, C. E. (1997). Trace-element contami-nation of a shallowwetland inWestern Australia.Marine andFreshwater Research, 48, 531–539.

Babica, P., Blaha, L., &Marsalek, B. (2006). Exploring the naturalrole of microcystins—a review of effects on photoautotro-phic organisms. Journal of Phycology, 42, 9–20.

Baldia, S., Evangelista, A., Aralar, E., & Santiago, A. (2007).Nitrogen and phosphorus utilization in the cyanobacteriumMicrocystis aeruginosa isolated from Laguna de Bay,Philippines. Journal of Applied Phycology, 19, 607–613.

Berry, J. P., Gantar, M., Perez, M. H., Berry, G., & Noriega, F. G.(2008). Cyanobacterial toxins as allelochemicals with poten-tial applications as algaecides, herbicides and insecticides.Marine Drugs, 6, 117–146.

Beutler, M., Wiltshire, K. H., Meyer, B., Moldaenke, C., Luring,C., Meyerhofer, M., Hansen, U. P., & Dau, H. (2002). Afluorometric method for the differentiation of algal popula-tions in vivo and in situ. Photosynthesis Research, 72, 39–53.

Beversdorf, L. J. (2013). Spatial and temporal variation incyanobacterial population dynamics and microcystin produc-tion in eutrophic lakes. Ph.D., The University of Wisconsin -Madison, Ann Arbor, USA.

Bittencourt-Oliveira, M.d., Chia, M., de Oliveira, H., CordeiroAraújo, M., Molica, R. & Dias, C. (2014). Allelopathicinteractions between microcystin-producing and non-

microcystin-producing cyanobacteria and green microalgae:implications for microcystins production. Journal of AppliedPhycology, 1–10.

Camacho, F.A. (2008). Algal chemical ecology. In C.D. Amsler(Ed.), (pp. 105–120) Heidelberg.

Chorus, I., & Bartram, J. (1999). Toxic cyanobacteria in water: aguide to their public health consequences, monitoring andmanagement. London: E & FN Spon.

Churro, C., Dias, E., & Valério, E. (2012). In Y. Lou (Ed.), Novelapproaches and their applications in risk assessment.Shanghai: InTech.

Coops, H., Beklioglu, M., & Crisman, T. (2003). The role ofwater-level fluctuations in shallow lake ecosystems – work-shop conclusions. Hydrobiologia, 506–509, 23–27.

Dadheech, P., Selmeczy, G., Vasas, G., Padisák, J., Arp, W.,Tapolczai, K., Casper, P., & Krienitz, L. (2014). Presenceof potential toxin-producing cyanobacteria in an oligo-mesotrophic lake in Baltic Lake District, Germany: anecological, genetic and toxicological survey. Toxins, 6,2912–2931.

Dai, G., Deblois, C. P., Liu, S., Juneau, P., & Qiu, B. (2008).Differential sensitivity of five cyanobacterial strains to am-monium toxicity and its inhibitory mechanism on the photo-synthesis of rice-field cyanobacterium Ge-Xian-Mi (Nostoc).Aquatic Toxicology, 89, 113–121.

Davis, J.A., Rosich, R.S., Bradley, J.S., Growns, J.E., Schmidt,L.G. & Cheal, F. (1993). Wetland classification on the basisof water quality and invertebrate community data.

Engström-Öst, J., Repka, S., & Mikkonen, M. (2011). Interactionsbetween plankton and cyanobacterium Anabaena with focuson salinity, growth and toxin production.Harmful Algae, 10,530–535.

Ghadouani, A., & Smith, R. E. H. (2005). Phytoplankton distri-bution in Lake Erie as assessed by a new in situ spectroflu-orometric technique. Journal of Great Lakes Research, 31,154–167.

Graneli, E., & Hansen, P. J. (2006). In E. Granéli & J. Turner(Eds.), Ecology of harmful algae (pp. 189–201). Heidelberg:Springer Berlin.

Graneli, E., Weberg, M., & Salomon, P. S. (2008). Harmful algalblooms of allelopathic microalgal species: The role of eutro-phication. Harmful Algae, 8, 94–102.

Hillebrand, H., Durselen, C., Kirschtel, D., Pollingher, U., &Zohary, T. (1999). Biovolume calculation for pelagic andbenthic microalgae. Journal of Phycology, 35, 403–424.

Holland, A., & Kinnear, S. (2013). Interpreting the possible eco-logical role(s) of cyanotoxins: compounds for competitiveadvantage and/or physiological aide? Marine Drugs, 11,2239–2258.

Huisman, J., & Hulot, F. D. (2005). In J. Huisman, H. C. P.Matthijs, & P. M. Visser (Eds.), Harmful cyanobacteria(pp. 143–176). Netherlands: Springer.

Huisman, J., Matthijs, H. C. P., Visser, P., Edwin, W., Kardinaal,A., & Visser, P. (2005). In J. Huisman, H. C. P. Matthijs, & P.M. Visser (Eds.), Harmful cyanobacteria (pp. 41–63).Netherlands: Springer.

ISO. (2005). Determination of microcystins: method using solidphase extraction (SPE) and high performance liquid chroma-tography (HPLC) with ultraviolet (UV) detection.

Istvánovics, V., Somlyódy, L., & Clement, A. (2002).Cyanobacteria-mediated internal eutrophication in shallow


Lake Balaton after load reduction. Water Research, 36,3314–3322.

Jang, M. H., Ha, K., Jung, J. M., Lee, Y. J., & Takamura, N.(2006). Increased microcystin production of Microcystisaeruginosa by indirect exposure of nontoxic cyanobacteria:potential role in the development of Microcystis bloom.Bulletin of Environmental Contamination and Toxicology,76, 957–962.

Janse, I., Kardinaal, W. E. A., Kamst-van Agterveld, M., Meima,M., Visser, P. M., & Zwart, G. (2005). Contrastingmicrocystin production and cyanobacterial population dy-namics in two Planktothrix-dominated freshwater lakes.Environmental Microbiology, 7, 1514–1524.

Jensen, J. P., Jeppesen, E., Olrik, K., & Kristensen, P. (1994).Impact of nutrients and physical factors on the shift fromcyanobacterial to chlorophyte dominance in shallow Danishlakes. Canadian Journal of Fisheries and Aquatic Sciences,51, 1692–1699.

Jeppesen, E., Brucet, S., Naselli-Flores, L., Papastergiadou, E.,Stefanidis, K., Nõges, T., Nõges, P., Attayde, J., Zohary, T.,Coppens, J., Bucak, T., Menezes, R., Freitas, F., Kernan, M.,Søndergaard, M., & Beklioğlu, M. (2015). Ecological im-pacts of global warming and water abstraction on lakes andreservoirs due to changes in water level and related changesin salinity. Hydrobiologia, 750, 201–227.

Jiang, Y., Ji, B., Wong, R. N. S., &Wong,M. H. (2008). Statisticalstudy on the effects of environmental factors on the growthand microcystins production of bloom-forming cyanobacte-rium Microcystis aeruginosa. Harmful Algae, 7, 127–136.

Joung, S.-H., Oh, H.-M., Ko, S.-R., & Ahn, C.-Y. (2011).Correlations between environmental factors and toxic andnon-toxic Microcystis dynamics during bloom in DaechungReservoir, Korea. Harmful Algae, 10, 188–193.

Kardinaal, W. E. A., Janse, I., Kamst-van Agterveld, M., Meima,M., Snoek, J., Mur, L. R., Huisman, J., Zwart, G., &Visser, P.M. (2007). Microcystis genotype succession in relation tomicrocystin concentrations in freshwater lakes. AquaticMicrobial Ecology, 48, 1–12.

Kearns, K. D., & Hunter, M. D. (2000). Green algal extracellularproducts regulate antialgal toxin production in a cyanobacte-rium. Environmental Microbiology, 2, 291–297.

Lawton, L. A., Edwards, C., & Codd, G. A. (1994). Extraction andhigh-performance liquid chromatographic method for thedetermination of microcystins in raw and treated waters.Analyst, 119, 1525–1530.

Legendre, P., & Legendre, L. (1998). Numerical ecology.Amsterdam: Elsevier.

Li, Y., & Li, D. (2011). Competition between toxic Microcystisaeruginosa and non-toxic Microcystis Wesenbergii withAnabaena PCC7120. Journal of Applied Phycology, 24,69–78.

Macías, F., Galindo, J., García-Díaz, M., & Galindo, J. (2008).Allelopathic agents from aquatic ecosystems: potential bio-pesticides models. Phytochemistry Reviews, 7, 155–178.

Millie, D. F., Fahnenstiel, G. L., Bressie, J. D., Pigg, R. J., Rediske,R. R., Klarer, D. M., Tester, P. A., & Litaker, R. W. (2009).Late-summer phytoplankton in western Lake Erie(Laurentian Great Lakes): Bloom distributions, toxicity, andenvironmental influences. Aquatic Ecology, 43, 915–934.

Nagai, T., Imai, A., Matsushige, K., & Fukushima, T. (2007).Growth characteristics and growth modeling of Microcystis

aeruginosa and Planktothrix agardhii under iron limitation.Limnology, 8, 261–270.

O’Neil, J. M., Davis, T. W., Burford, M. A., & Gobler, C. J.(2012). The rise of harmful cyanobacteria blooms: the poten-tial roles of eutrophication and climate change. HarmfulAlgae, 14, 313–334.

Orihel, D. M., Bird, D. F., Brylinsky, M., Chen, H. R., Donald, D.B., Huang, D. Y., Giani, A., Kinniburgh, D., Kling, H.,Kotak, B. G., Leavitt, P. R., Nielsen, C. C., Reedyk, S.,Rooney, R. C., Watson, S. B., Zurawell, R. W., &Vinebrooke, R. D. (2012). High microcystin concentrationsoccur only at low nitrogen-to-phosphorus ratios in nutrient-rich Canadian lakes. Canadian Journal of Fisheries andAquatic Sciences, 69, 1457–1462.

Paerl, H. W., & Otten, T. G. (2013). Harmful cyanobacterialblooms: causes, consequences, and controls. MicrobialEcology, 65, 995–1010.

Pflugmacher, S. (2002). Possible allelopathic effects ofcyanotoxins, with reference to microcystin-LR, in aquaticecosystems. Environmental Toxicology, 17, 407–413.

Pinheiro, C., Azevedo, J., Campos, A., Loureiro, S., &Vasconcelos, V. (2013). Absence of negative allelopathiceffects of cylindrospermopsin and microcystin-LR on select-ed marine and freshwater phytoplankton species.Hydrobiologia, 705, 27–42.

Reichwaldt, E. S., Song, H., & Ghadouani, A. (2013). Effectsof the distribution of a toxic Microcystis bloom on thesmall scale patchiness of zooplankton. Plos One, 8,e66674.

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

Reynolds, C.S., Usher, M., Saunders, D., Dobson, A., Peet, R.,Adam, P., Birks, H.J.B., Gustafssor, L., McNelly, J., Paine,R.T. & Richardson, D. (2006). The ecology of phytoplankton,(pp. 178–238). Cambridge University Press.

Rinta-Kanto, J. M., Konopko, E. A., DeBruyn, J. M.,Bourbonniere, R. A., Boyer, G. L., & Wilhelm, S. W.(2009). Lake Erie Microcystis: relationship betweenmicrocystin production, dynamics of genotypes and environ-mental parameters in a large lake. Harmful Algae, 8, 665–673.

Rzymski, P., Poniedziałek, B., Kokociński,M., Jurczak, T., Lipski,D., & Wiktorowicz, K. (2014). Interspecific allelopathy incyanobacteria: Cylindrospermopsin and Cylindrospermopsisraciborskii effect on the growth and metabolism ofMicrocystis aeruginosa. Harmful Algae, 35, 1–8.

Scott, L. L., Downing, S., Phelan, R. R., &Downing, T. G. (2014).Environmental modulation of microcystin and β-N-methylamino-l-alanine as a function of nitrogen availability.Toxicon, 87, 1–5.

Sinang, S., Reichwaldt, E., & Ghadouani, A. (2013). Spatial andtemporal variability in the relationship betweencyanobacterial biomass and microcystins. EnvironmentalMonitoring and Assessment, 185, 6379–6395.

Soares, M. C. S., Rocha, M. I. A., Marinho, M. M., Azevedo, S.M. F. O., Branco, C. W. C., & Huszar, V. L. M. (2009).Changes in species composi t ion during annualcyanobacterial dominance in a tropical reservoir: physicalfactors, nutrients and grazing effects. Aquatic MicrobialEcology, 57, 137–149.


Švercel, M. (2013). Cyanobacteria: ecology, toxicology and man-agement. In S. Aloysio Da Ferrão-Filho (Ed.), (pp. 27–41),Nova Science Publishers, United States of America.

Utermöhl, H. (1958). Zur vervollkommnung der quantitativenphytoplankton-methodik. Mitteilungen InternationaleVereinigung fur Theoretische und Angewandte Limnologie,9, 1–38.

Utkilen, H., & Gjolme, N. (1995). Iron-stimulated toxin produc-tion in Microcystis aeruginosa. Applied and EnvironmentalMicrobiology, 61, 797–800.

Van de Waal, D. B., Verspagen, J. M. H., Lürling, M., Van Donk,E., Visser, P. M., & Huisman, J. (2009). The ecologicalstoichiometry of toxins produced by harmful cyanobacteria:an experimental test of the carbon-nutrient balance hypothe-sis. Ecology Letters, 12, 1326–1335.

Vezie, C., Rapala, J., Vaitomaa, J., Seitsonen, J., & Sivonen, K.(2002). Effect of nitrogen and phosphorus on growth of toxicand nontoxic Microcystis strains and on intracellularmicrocystin concentrations.Microbial Ecology, 43, 443–454.

Woodward, B. (2008). Literature and Interview Project:Constructed Lakes in the Perth Metropolitan and SouthWest Region. Prepared for Department of Water, WesternAustralian Local Government Association.

Wu, Y., Li, L., Gan, N., Zheng, L., Ma, H., Shan, K., Liu, J., Xiao,B., & Song, L. (2014). Seasonal dynamics of water bloom-forming Microcystis morphospecies and the associated ex-tracellular microcystin concentrations in large, shallow, eu-trophic Dianchi Lake. Journal of Environmental Sciences,26, 1921–1929.

Yang, J., Deng, X., Xian, Q., Qian, X., & Li, A. (2014).Allelopathic effect of Microcystis aeruginosa onMicrocystis wesenbergii: microcystin-LR as a potentialallelochemical. Hydrobiologia, 727, 65–73.

Zurawell, R.W., Chen, H. R., Burke, J.M., & Prepas, E. E. (2005).Hepatotoxic cyanobacteria: a review of the biological impor-tance of microcystins in freshwater environments. Journal ofToxicology and Environmental Health-Part B-CriticalReviews, 8, 1–37.


Could the presence of larger fractions of non-cyanobacterial species be used as a predictor of...

Documents

Transcript of Could the presence of larger fractions of non-cyanobacterial species be used as a predictor of...