Effects of the pyrethroid insecticide, cypermethrin, on afreshwater community studied under field conditions. I.Direct and indirect effects on abundance measures of

organisms at different trophic levels

Ursula Friberg-Jensen a,*, Lina Wendt-Rasch b, Per Woin b,Kirsten Christoffersen a

a Freshwater Biological Laboratory, University of Copenhagen, Helsingørsgade 51, DK-3400 Hillerød, Denmarkb Department of Ecology, Chemical Ecology and Ecotoxicology, University of Lund, Ecology Building, Solvegatan 37, S-22362 Lund,

Sweden

Received 15 February 2002; received in revised form 18 November 2002; accepted 19 November 2002

Abstract

The effects of the pyrethroid insecticide cypermethrin on a natural freshwater community were studied in small in situenclosures over an 11-day period. The experiment was conducted in a eutrophic lake using a regression design thatincluded three untreated controls and a gradient of six unreplicated cypermethrin concentrations, ranging from 0.01 to6.1 mg/l. This paper is the first in a series of two, and describes the fate of cypermethrin and its effects on the abundanceof crustaceans, rotifers, protozoans (cilliates and heterotrophic nanoflagellates (HNF)) and bacteria and the biomass ofperiphytic and planktonic algae. The concentration of cypermethrin decreased quickly during the experiment, with ahalf-life of 48 h for the total and 25 h for the dissolved fractions of cypermethrin, respectively. Cypermethrin proved tobe acutely toxic to crustaceans in enclosures receiving nominal cypermethrin concentrations of ]/0.13 mg/l. No EffectConcentration (NEC) and median Effect Concentration (EC50) for the total crustacean community and cladoceran andcopepod subgroups ranged between 0.02!/0.07 and 0.04!/0.17 mg/l, respectively, with copepods being less sensitive thancladocerans. The abundance of rotifers, protozoans and bacteria and the chlorophyll-a concentration of planktonic andperiphytic algae was significantly related to the concentration of cypermethrin. All groups proliferated within 2!/7 daysafter the cypermethrin application in those enclosures where the abundance of crustaceans was seriously affected bycypermethrin (i.e. ]/0.13 mg/l). We hypothesise that the proliferation of rotifers, protozoans, bacteria and algae was dueto a reduced grazer control from crustaceans and thereby mediated indirectly by cypermethrin. The results of thisexperiment provide knowledge on how an entire microplankton community may respond to pyrethroids in nature, andthe indirect effects observed on the community clearly demonstrates the necessity of multispecies field experiments inecotoxicological risk assessment.# 2003 Elsevier Science B.V. All rights reserved.

* Corresponding author. Tel.: "/45-48267600; fax: "/45-4824-1476.E-mail address: urfribergjensen@zi.ku.dk (U. Friberg-Jensen).

Aquatic Toxicology 63 (2003) 357!/371

www.elsevier.com/locate/aquatox

0166-445X/02/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.doi:10.1016/S0166-445X(02)00201-1

Keywords: Pyrethroids; No effect concentration; Crustaceans; Protozoans; Algae; Indirect effects; Ecosystem effects

1. Introduction

Many lakes, especially shallow lakes in agricul-tural areas, have changed from macrophyte domi-nated clearwater stages into algal dominatedturbid ecosystems during the last decades (Phillipset al., 1978; Hough et al., 1989). Such shifts arethought to arise from eutrophication and succeed-ing algal blooms. In addition to nutrients, pesti-cides have been suggested as a possible factorinducing alga blooms (Hurlbert et al., 1971; Brockand Budde, 1994). Several recent studies show thatpesticide residues frequently occur in surfacewaters (Anonymous, 1999; Larson et al., 1999;Ulen et al., 2002) dispersed from sprayed agricul-tural fields through, e.g. spray-drift and runoff.

The insecticide cypermethrin is one of severalsynthetic pyrethroids used worldwide to controlinsect pests. Cypermethrin was introduced on themarket in the late 1970s and has since been usedon a wide range of crops due to its high pesticidalactivity and low mammalian toxicity relative toother insecticides (Davies, 1985). However, severalstudies have shown that aquatic invertebrates andfish are extremely sensitive to cypermethrin andpyrethroids in general (Anderson, 1989; Day,1989; Haya, 1989; Mian and Mulla, 1992). Directtoxic effects on crustacean zooplankton includereduced feeding activity (Day and Kaushik,1987a), reproduction and survival (Day andKaushik, 1987b; Kjølholt et al., 1991).

The crustacean zooplankton constitute an im-portant component of freshwater ecosystems, aseffective grazers on most planktonic microparti-cles. In mesotrophic and eutrophic freshwaters thecrustacean zooplankton can depress the biomassof phytoplankton, protozoans and rotifers effi-ciently; at least in the absence of planktivorous fish(Lampert et al., 1986; Gulati, 1990; Christoffersenet al., 1993). Consequently, direct toxic effects onthe crustacean zooplankton may result in areduced grazer control of phytoplankton, proto-zoans and rotifers, which may lead to an alteredbiomass and species composition of the commu-

nities. Thus, responses to pyrethroids may involveboth direct effects on crustacean zooplanktoncommunities and indirect cascading effects onother trophic levels in the aquatic foodweb. Thiscan clearly affect the structure of the ecosystem ina way that may resemble, as well as reinforce,changes caused by eutrophication.To our knowledge no field studies have focused

on the effects of pyrethroid insecticides on entiremicrobial plankton communities. Today’s riskassessment of pesticides in aquatic environmentsis largely based on results from single-specieslaboratory studies using standard test speciesrepresenting primary- and secondary producersand carnivores in the classical grazer food web(European Commission, 1995). However, thesetest procedures provide no opportunity to studyany indirect effects of pesticides caused by altera-tions in biotic interactions like predation andcompetition. A more holistic view of the potentialenvironmental hazards of a chemical requiresmultispecies test systems, which allow the occur-rence of such biotic interactions.In this field study we, therefore, aimed to obtain

a better understanding of the effects of pyrethroidson freshwater ecosystems by examining the impactof cypermethrin on several organisms includingcrustaceans, rotifers, protozoans (heterotrophicnanoflagellates (HNF) and ciliates), bacteria andperiphytic and planktonic algae following a singleexposure to cypermethrin. The experimental setupconsisted of small in situ enclosures and was basedon an unreplicated regression design earlier de-scribed by Liber et al. (1992), which permitted theestimation of No Effect Concentrations (NEC)and median Effect Concentrations (EC50) ofexposed organisms. This paper is the first in aseries of two, and focuses on the fate of cyperme-thrin and on the direct and indirect effects ofcypermethrin on the abundance of planktonicorganisms at different trophic levels. The secondpaper focuses on the effects of cypermethrin on thespecies composition within different trophic levels(Wendt-Rasch et al., 2002).

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371358

2. Materials and methods

2.1. Experimental design

The experiment was conducted in nine pelagicenclosures (0.1 mm polyethylene; width 44 cm;depth 150 cm) filled with 200 l surface water fromthe eutrophic Lake Frederiksborg Slotssø, Den-mark, in June 1999. The enclosures were attachedto three wooden frames that were anchored 50 mfrom the shore. To ensure that all epilimneticzooplankton genera were present in the enclosureshalf a litre of water containing zooplankton fromten pooled plankton net (140 mm) hauls was addedto each enclosure.

The treatment concentrations of cypermethrin(Cyano-(3-phenoxyphenyl)-methyl-3-(2,2-di-methylcyclopropancarboxylat); Riedel-deHaen;96% cypermethrin; analytical standard grade)were prepared by diluting the solid crystals inpure acetone. Six enclosures were treated withcypermethrin 22 h after the establishment of theenclosures; the remaining three enclosures servedas replicated controls. Cypermethrin was added tothe water 30 cm below the surface and mixed intothe water column by gently stirring for 2 mingiving nominal initial concentrations of 0.01, 0.04,0.13, 0.47, 1.7 and 6.1 mg cypermethrin/l. Theconcentration of acetone in the six cypermethrintreated enclosures did not exceed 0.24 ml acetone/l,which is well below the limit of 0.1 ml/l recom-mended by the International Standard Organiza-tion (1996) and was consequently not tested fortoxicity. No acetone was added to the controlenclosures.

2.2. Physical/chemical parameters andcypermethrin analysis

Oxygen, conductivity and temperature weremeasured with a HYDROLAB minisonde andpH was measured with a pH-meter (PHM 85,Radiometer, Copenhagen) on each sampling date.

Samples for cypermethrin analysis were taken inall nine enclosures 3 h after the cypermethrinapplication, and subsequently on day 1 in theenclosures holding nominal concentrations of0.13!/6.1 mg/l, on day 4 in the enclosures holding

nominal concentrations of 0.47!/6.1 mg/l, andfinally on day 11 in the enclosure holding thehighest concentration (6.1 mg/l). Sampling wasconducted using solid phase extraction (SPE) onsite after modifying the method by Woin (1994).The water was sucked directly from the enclosureat a depth of 10 cm by a hand operated vacuumequipment through a glass micro fibre filter (GF/F) connected to the front of the SPE column(INSOLUTE† ENV"/from International SorbentTechnology). Extracted water was collected in avacuum flask and the volume was determined(500!/1000 ml). Glass fibre filters and SPE col-umns were transferred to the laboratory andstored in a freezer (#/20 8C) for analysis ofparticulate bound and dissolved fractions ofcypermethrin. Due to low expected recovery ofcypermethrin on the two last sampling occasions(in the enclosure with the nominal concentrationof 1.7 mg/l on day 4 and in the enclosure with thenominal concentration of 6.1 mg/l on day 11) thesamples was not separated in two fractions (parti-culate and dissolved) prior to analysis. The filterswere extracted with a mixture of dichloromethaneand acetone (1:1, v/v) in a Soxtec Avanti 2050automatic extraction system. Analytes were elutedfrom the SPE columns using 2$/3 ml of dichlor-omethane. Extracts were analysed on a Hewlett!/

Packard model 5890 gas chromatograph equippedwith two 63Ni electron-capture detectors andtwo columns (CP-Sil 19 CB and CP-Sil 5 CBwith dimensions of 20 m$/0.32 mm i.d. and 0.25mm film thickness provided by ChrompackSverige AB, Nacka, Sweden) attached to thesame injector. The injection volume was 2 ml andthe injection was splitless. Injector temperaturewas 250 8C and detector temperature was 300 8C.Oven temperature was set to 90 8C for 1 min,increasing 30 8C/min to 180 8C and then 4 8C/min to 260 8C, where it was held for 12 min.Soxtec extraction and gas chromatographic deter-mination was performed at Department of Envir-onmental Assessment, Sweden University ofAgricultural Sciences. Analytical standards wereprovided by Dr Ehrenstorfer Gmbh, Augsburg,Germany.

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371 359

2.3. Planktonic communities

Water samples were taken with a cylindricalwater sampler (width 3.5 cm; length 120 cm) in allenclosures before the application of cypermethrin(day 0) and subsequently 4, 12 h and 1, 2, 4, 7 and11 days after the application. A total of five watersamples evenly distributed within each enclosurewere taken on each sampling occasion. Threesamplers were used: one for the controls, one forthe three lowest and one for three highest con-centrations of cypermethrin. The samplers werecarefully rinsed in lake water between each sam-pling. Water samples from each enclosure werefurther subsampled for analysis of the abundanceof crustaceans and rotifers after 4 and 12 h and onday 0, 1, 2, 4, 7 and 11 and the abundance ofciliates, HNF, bacteria and the phytoplanktonbiomass on day 0, 1, 2, 4, 7 and 11.

For counting and identification of crustaceansand rotifers, a 1 l subsample was filtered through a200 mm mesh net (mesozooplankton) and subse-quently through a 50 mm mesh net (rotifers andnauplii). The retained animals were fixed withLugol’s solution. Enumeration and identificationto species or genus level was done at 100!/400$/

magnification using an inverted microscope(Olympus IMT-2).

The numbers of ciliates were counted in aninverted microscope (Olympus IMT-2) at 300$/

magnification after sedimentation of an unfiltered5 ml subsample fixed with Lugol’s solution.

The number of HNF was determined using theDAPI staining technique (Porter and Feig, 1980).Water samples fixed with glutaraldehyde (finalconcentration 2%) were stained with DAPI (4?5-diamidino-2-phenylindol) to a concentration of 0.5mg DAPI per ml for 20 min. Each sample (1 ml)was filtered through a 0.8 mm black polycarbonatefilter and mounted on a glass slide with immersionoil. The number of HNF was counted in anepifluorescence microscope (Olympus BH-2) at1250$/ magnification.

Bacteria were stained and counted as describedfor HNF with the exceptions that the finalconcentration of DAPI was 0.1 mg/ml and thatthe samples were filtered through 0.2 mm blackpolycarbonate filters.

Phytoplankton biomass was measured by spec-trophotometry as the concentration of chloro-phyll-a in 0.5-1.0 l GF/C filtrated water samplesafter extraction with ethanol, according to themethod described by Jespersen and Christoffersen(1987).

2.4. Periphyton

Periphyton was sampled from unglazed ceramictiles (3.7 cm2) that served as artificial substrates.Seven tiles were placed in a plastic (HD polythene)frame and allowed to pre-colonise with a periphy-tic community for 10 days in the eutrophic LakeBysjon (Scania, Sweden). At the start of theexperiment (day 0), two frames with pre-colonisedtiles were positioned 10 cm below the water surfacein each of the enclosures. At the end of theexperiment (day 11) seven ceramic tiles fromeach enclosure were allowed to dry in darkness(2 h) and subsequently frozen (#/80 8C) for laterchlorophyll-a analyses using the method describedby Jespersen and Christoffersen (1987).

3. Data analysis

3.1. Fate of cypermethrin

By way of linear regression, actual concentra-tions of total and dissolved cypermethrin werecalculated for the time t%/0. Linear regressionswere performed on log-transformed concentrationmeasures expressed as the proportion of therespective nominal concentrations (measurementscorresponding to the nominal concentrations 0.01and 0.04 mg/l were not included because they wereclose to the detection limit). The actual totalconcentration of cypermethrin (i.e. particle boundplus dissolved fraction) in each enclosure at thetime t%/0 were used in the estimation of NEC andEC50 for the crustacean zooplankton.The half-life for the total (i.e. particle bound

plus dissolved fraction) and dissolved fractions ofcypermethrin was calculated by solving the equa-tion of the linear regression. All calculations werebased on the assumption that the same degrada-

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371360

tion processes were acting in all enclosures irre-spective of the initial concentration.

3.2. Crustacean zooplankton

The crustacean density was 50!/75% lower in allpre-exposure samples including the controls, thanin samples taken 4 h after the application ofcypermethrin. This was probably due to a timelag effect between the experimental setup and thefirst sampling occasion (B/1 h), resulting in theorganisms displaying some kind of escape beha-viour like staying near the sides of the enclosurewalls, as a response to the experienced waterturbulence during the experimental setup. As aconsequence the average density in the controls onthe sampling time 4 h after cypermethrin applica-tion was used as the start (day%/0) density in allenclosures.

NEC and EC50 were calculated for the predo-minant species or genera of crustaceans (i.e. anaverage of ]/35 ind per l in the controls at thegiven day) by way of inverse regression analysis.Data were plotted as log dose!/log response curveswith a separate line for each sampling occasion.Inverse regression analysis was performed on thelinear part of the log-transformed dose!/responsedata sets to obtain NEC- and EC50-values inaccordance with Liber et al. (1992). NEC- andEC50-values were calculated from the equation forthe regression line as:

log(Y )%a"b log(X )

where Y is the response value (abundance) in thecontrol treatment to the given sampling time, a isthe intercept, b is the slope of the regression line,and X is the NEC or EC50-value.

The 95 % confidence limits of the NEC-and EC50-values were calculated by use of

the following equation:

where X/%/mean of the logarithm to the includedconcentrations, Y0%/upper/lower 95% confidencelimit for the control value, Y/%/mean of thelogarithm to the Y -value corresponding to theconcentration X ; b%/slope of the regression line,t%/t -value corresponding to the actual degrees offreedom (n-1), s%/sample standard deviation,Sxx%S(Xi#X )2; Xi%/logarithm to the concen-trations included in the regression, n%/number ofdata points in the regression analysis.A detailed examination of the statistical princi-

ples in the regression design and the estimation of95% confidence intervals are given by Draper andSmith (1981) and Liber et al. (1992).In contrast to Liber et al. (1992) who used

the average abundance in all enclosures on day 0as the control value during the entire study periodwe used the average abundance in the controlenclosures at the given day in the calculationof NEC and EC50. The rationale behind this isthat most planktonic organisms have very shortgeneration times and their abundances oftenfluctuate very rapidly in response to predationand food availability. Using the same controlvalue during the entire experimental period might,therefore, be incorrect when studying planktoniccommunities.

3.3. Rotifers, protozoans, bacteria and algae

Simple linear regression was used to evaluate therelationship between cypermethrin exposure andthe abundance of rotifers, ciliates, HNF, bacteriaand the amount of planktonic and periphyticchlorophyll-a on each sampling date. The regres-sions were performed on log-transformed data inorder to meet the requirements of homogeneity ofvariance and normalised statistical distributions.

NEC or ECupper

NEC or EClower

g%X"b(Y0 # Y )9 tsf[(Y0 # Y )2=Sxx]" (b2=n)# (t2s2=nSxx)g

1=2

b2 # (t2s2=Sxx)

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371 361

4. Results

4.1. Chemical and physical parameters

The nine individual enclosures had similarvalues for temperature, oxygen content, conduc-tivity and pH throughout the experiment. Thetemperature varied between 16.3 and 19.0 8C anddissolved oxygen, conductivity and pH variedbetween 10.5 and 12.4 mg O2 per l, 484!/502 mS/cm2 and 8.3!/9.1, respectively.

The actual concentration of cypermethrin de-creased in all the treated enclosures during theexperiment (Table 1). Linear regression on log-transformed concentrations revealed that the total(i.e. particulate bound plus dissolved fraction) anddissolved fractions of cypermethrin in the watercolumn decreased significantly with time (r2%/

0.92, PB/0.001 and r2%/0.88, PB/0.01). The rateof disappearance followed a first order kineticmodel and the half-lives for the total and dissolvedfractions of cypermethrin were 48 and 25 h,respectively. The largest amount of recoveredcypermethrin was bound to particles, as 56!/

100% of the analyte was collected in the GF/Ffilter fraction. The highest measured concentra-tions of the total (3.2 mg/l) and dissolved (0.8 mg/l)fractions of cypermethrin were found in theenclosure with the highest nominal concentration(6.1 mg/l) 3 h after the application of cypermethrin.Dissolved fractions of cypermethrin were never

recorded in enclosures treated with nominal con-centrations below 0.47 mg/l. No cypermethrin wasfound in the control enclosures.

4.2. Effects on crustacean zooplankton

Following the application of cypermethrin thetotal abundance of crustaceans (dominated by oneDaphnia species and one calanoid copepod spe-cies) decreased rapidly in the three high treatmentenclosures (0.47!/6.1 mg/l) in contrast to thecontrol and low treatment enclosures (Fig. 1A).Four hours after the application the total abun-dance in the three high treatment enclosurescorresponded to 31, 18 and 5% of the controland after 12 h the abundance of crustaceans in theenclosure treated with 0.13 mg/l was also reducedto 34% of the control. On day 2 no crustaceanswere present in the highest concentration. Afterthe fourth day, the total abundance of crustaceansincreased in the enclosures with nominal concen-trations of 0.47 and 1.7 mg cypermethrin/l althoughit never reached the control level (Fig. 1A). Theserises were attributed to a gradual increase in theabundance of copepod nauplii (Fig. 1C), as out-lined by Wendt-Rasch et al. (2002). From 4 h today 4 after the initial exposure the NEC and EC50

based on actual estimated cypermethrin concen-trations for the total abundance of crustaceans andthe cladoceran and copepod subgroups rangedbetween 0.02 and 0.07 mg cypermethrin/l and 0.04!/

Table 1Nominal, calculated and measured concentrations of cypermethrin (mg/l) in the enclosures

Enclosurea Startb 3 h Day 1 Day 4 Day 11

T P DS T P DS T P DS T P DS T

Control 0 0 0 0 n.a n.a n.a n.a n.a n.a n.a n.a n.a0.01 0.006 0.10 n.d 0.10 n.a n.a n.a n.a n.a n.a n.a n.a n.a0.04 0.024 0.08 n.d 0.08 n.a n.a n.a n.a n.a n.a n.a n.a n.a0.13 0.078 0.08 n.d 0.08 n.d n.d n.d n.a n.a n.a n.a n.a n.a0.47 0.281 0.20 0.10 0.30 0.10 n.d 0.1 n.d n.d n.d n.a n.a n.a1.70 1.015 0.90 0.70 1.6 0.40 0.20 0.60 n.a n.a 0.4 n.a n.a n.a6.10 3.643 2.40 0.8 3.20 2.20 0.60 2.80 0.60 0.10 0.70 n.a n.a 0.08

T, sum of particle bound and dissolved fractions; P, particle bound fraction; DS, dissolved fraction; n.a, not analysed; n.d, nodetectable concentration.

a Nominal concentrations (mg/l).b Calculated start concentrations (t%/0).

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371362

0.17 mg cypermethrin/l, respectively (Table 2). TheNEC and EC50-values based on nominal cyperme-thrin concentrations ranged between 0.03 and 0.11mg cypermethrin/l and 0.04!/0.29 mg cypermethrin/l, respectively. NEC and EC50-values for the totalzooplankton community and cladocerans at 4 hwas higher than the respective values at later times(Table 2) and the maximum impact of the insecti-

cide took place after 12 h!/4 days of exposure (Fig.1). Copepods were generally more sensitive thancladocerans in the first 12 h of exposure and lesssensitive than cladocerans from day 2 (Table 2 andFig. 1).

4.3. Effects on rotifers, protozoans and bacteria

The abundance of rotifers increased from day 4in all enclosures treated with ]/0.13 mg cyperme-thrin/l and from day 7 in the low treatmentenclosures (0.01!/0.04 mg cypermethrin/l) (Fig.2A). No increase in the abundance of rotiferswas observed in the control enclosures. In theenclosure with the highest nominal cypermethrinconcentration the abundance had increased dra-matically by a factor 17 at the end of theexperiment while it had only doubled in the lowesttreatment compared with the control. Linearregression showed that the abundance of rotiferswas significantly related to the concentration ofcypermethrin on day 4 (r2%/0.45), day 7 (r2%/0.77)and day 11 (r2%/0.86) (Fig. 3A).The abundance of ciliates was constant in the

control enclosures and the three low treatmentenclosures until day 2, after which the abundancedeclined. The ciliate abundance in the enclosureswith the three highest concentrations (0.47!/6.1 mg/l) increased initially by a factor 2.5!/4 on day 2compared with the control and remained at ahigher level in these enclosures throughout the restof the study period (Fig. 2B). Linear regressionshowed that the abundance of ciliates was sig-nificantly related to the concentration of cyperme-thrin during the entire study period (r2 rangingbetween 0.52 and 0.90) (Fig. 3B).The abundance of HNF declined in a similar

pattern in the controls and in the two lowtreatment enclosures during the course of theexperiment (Fig. 2C). By contrast, the abundanceof HNF in all enclosures treated with ]/0.13 mgcypermethrin/l increased substantially comparedwith the level in the controls until day 7. Despite asubsequent decline in the abundance of HNF inthe four high treatment enclosures, the abundancewas still 2.5!/4 times higher than the control valuesat the end of the experiment. Linear regressionshowed that the abundance of HNF was signifi-

Fig. 1. The abundance of (A) total crustacean zooplankton, (B)cladocerans and (C) copepods in control and cypermethrintreated enclosures. Cypermethrin concentrations are nominalconcentrations. Error bars indicate the standard deviation ofthe controls (n%/3). Note that the y -axis is logarithmic scale.

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371 363

cantly related to the concentration of cyperme-thrin on day 2 (r2%/0.63), day 4 (r2%/0.86), day 7(r2%/0.87) and day 11 (r2%/0.79) (Fig. 3C).

The bacterial abundance declined initially in allenclosures. A modest regrowth started from day 2in the control and in the two lowest cypermethrintreatments whereas a more pronounced increasewas observed in the high treatment enclosures(0.47!/6.1 mg/l) (Fig. 2D). At the end of theexperiment (day 11) the bacterial abundance hadapproximately doubled in the three high treatmentenclosures compared with the controls. Linearregression showed that the abundance of bacteriawas significantly related to the concentration ofcypermethrin on day 7 (r2%/0.86) and day 11 (r2%/

0.92) (Fig. 3D).

4.4. Effects on phytoplankton and periphyton

The amount of phytoplankton, measured as theconcentration of chlorophyll-a decreased in allenclosures from day 4. However, in the enclosurestreated with 0.13!/6.1 mg/l the chlorophyll-a con-centration was up to twice the concentration in thecontrol and the two low treatment enclosures inthe period from day 2 to 11 (Fig. 2E). Linearregression showed that the concentration of chlor-

ophyll-a was significantly related to the concen-tration of cypermethrin on day 2 (r2%/0.54), day 4(r2%/0.75), day 7 (r2%/0.78) and day 11 (r2%/0.53)(Fig. 3E).The biomass of periphytic algae on the ceramic

tiles was higher in the enclosures exposed tocypermethrin compared with the controls andlinear regression revealed that the amounts ofchlorophyll-a (mg chl-a /cm2) were significantlyrelated to the cypermethrin concentration on day11 (PB/0.003, r2%/0.73) (Fig. 4).

5. Discussion

The calculated half-life of 48 h and 25 h for thedetectable total (i.e. particle bound plus dissolvedfraction) and dissolved amounts of cypermethrinin the enclosures was in accordance with pre-viously reported half-lives in the water column of 1(Farmer et al., 1995) to 2 days (Leahey, 1985). Asexpected, the largest amount of recovered cyper-methrin was bound to particles since cypermethrinis a highly hydrophobic chemical (log Kow%/6.6)that is adsorbed rapidly and binds strongly tosurfaces and organic matter (Sharom and Solo-mon, 1981; Kjølholt et al., 1991; Hill et al., 1994).

Table 2NEC and EC50 (mg/l) for the total crustacean community and cladoceran and copepod subgrups based on actual estimatedcypermethrin concentrations

Sample time NEC (mg/l) EC50 (mg/l) n r2

Total crustaceans 4 h 0.07 (0.013!/0.179) 0.17 (0.035-0.172) 4 0.9812 h 0.02 (0.004!/0.079) 0.04 (0.006!/*) 5 0.98Day 1 0.03 (0.013!/0.082) 0.06 (0.0203!/*) 4 0.99Day 2 0.06 (0.001!/0.351) 0.09 (0.002!/*) 4 0.94Day 4 0.02 (0.003!/0.097) 0.04 (0.005!/*) 5 0.96

Cladocerans 4 h 0.07 (0.002!/0.381) 0.17 (0.006!/*) 4 0.9512 h 0.04 (0.028!/0.121) 0.06 (0.034!/*) 3 0.99Day 1 0.04 (0.013!/0.162) 0.06 (0.017!/*) 4 0.99Day 2 0.03 (0.008!/*) 0.06 (0.010!/*) 4 0.99Day 4 0.02 (0.009!/0.057) 0.03 (0.012!/*) 4 0.99

Copepods 4 h 0.03 (0.002!/0.119) 0.11 (0.010!/*) 5 0.9212 h 0.02 (0.001!/0.099) 0.05 (0.003!/*) 5 0.94Day 1 0.03 (0.016!/0.063) 0.06 (0.027!/0.294) 4 0.99Day 2 0.06 (0.001!/0.246) 0.11 (0.003!/*) 4 0.94Day 4 0.06 (0.02!/0.151) 0.11 (0.027!/0.341) 4 0.98

Numbers in brackets are 95% confidence limits. n , The number of data points included in the regression analysis (n max%/6); r2,thecoefficient of correlation for the regression line, *, no value.

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371364

For the same reason the amount of cypermethrinthat was not recovered 3 h after the application(B/50% of the total added) may have beenadsorbed to the walls of the plastic bags and/orto particles that had settled down within the 3 h.Laboratory studies using sediment-water systemshave shown that once pyrethroids become ad-sorbed they are much less bioavailable and hencepotential effects are substantially reduced (Hill,1985). The reason for using the estimated concen-tration of cypermethrin instead of the nominalconcentration of cypermethrin in the calculation ofNEC and EC50 is that it more accurately reflectsthe concentrations to which the organisms wereexposed. This was particular true at the highestexposure concentration (6.1 mg/l) which was abovethe solubility level for cypermethrin in water and,therefore, adsorbed with rapidity to the particlesand the enclosure walls.Cypermethrin is a neurotoxic insecticide that

mainly acts as a disruptor of the kinetics of thesodium channels of nerve cells. This results inprolonged depolarisation of the nerve membranescausing convulsions, paralysation and eventuallydeath (Miller and Salgado, 1985; Soderlund andBloomquist, 1989; Bloomquist, 1996). The toxiceffect of cypermethrin on naturally living crusta-cean zooplankton was clearly demonstrated in thepresent field study, as their abundance decreasedsubstantially within just 4 h in all enclosureshaving nominal cypermethrin concentrationsequal to or higher than 0.13 mg/l (Fig. 1). Themain impact on the crustaceans did not occur untilafter 4 h of exposure, as illustrated by the NECand EC50-values that are generally higher 4 h afterthe application compared with later exposuretimes (Table 2). Copepods were generally lessaffected than cladocerans as they where able torecover in enclosures holding nominal cyperme-thrin concentrations equal to or lower than 0.47mg/l (Fig. 1B, C). This response was attributed todifferences in sensitivity within the copepod sub-group as outlined by Wendt-Rasch et al. (2002).The calculated NEC-values were in accordance tofield results by Farmer et al. (1995), Shires andBennett (1985) who found no effects on theabundance of crustaceans at actual measuredcypermethrin concentrations of around 0.03 mg/l.

Fig. 2. Abundance of (A) rotifers ($/102/l), (B) ciliates ($/103/l), (C) HNF ($/106/l), (D) bacteria ($/108/l) and (E) theconcentration of chlorophyll-a in control and cypermethrintreated enclosures. The cypermethrin concentrations are nom-inal concentrations. Error bars indicate the standard deviationof the controls (n%/3).

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371 365

It appears from the results that it was not possibleto determine NEC- and EC50-values or estimatethe associated 95% confidence limits at all sam-pling times and for all identified crustaceanspecies; see also Wendt-Rasch et al. (2002).Potential reasons for this include either a lack oflog linearity in the response to the cypermethrinconcentration gradient or a large variation in the

abundance of crustaceans in the control enclosuresleading to broad confidence limits for the controlvalues. In general, the 95% confidence limitsaround the NEC and EC50-values covered a verybroad range. Thus, EC50 values were in themajority of cases smaller than the upper confi-dence limit of NEC, which indicated that we wereunable to identify an effect as significant unless the

Fig. 3. The relation between log-transformed abundance of (A) rotifers, (B) ciliates, (C) HNF, (D) bacteria, (E) chlorophyll-aconcentration and the log-transformed nominal cypermethrin concentration on the different sampling days. *, Significant relation PB/

0.05; **, PB/0.01; ***, PB/0.001.

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371366

abundance was reduced by at least 50%. Similarproblems was experienced by Liber et al. (1992) intheir use of the regression design, as EC50 in nineout of 14 cases was lower than the upper con-fidence limit of NEC. One way to improve theprecision could be to increase the number ofincluded test concentrations and/or the numberof control enclosures.

In contrast to the crustaceans, no direct negativeeffect of the applied cypermethrin concentrationswas observed on any other of the examinedheterotropic or autotrophic organisms in thepresent study (Fig. 2). This is in accordance withour expectations, since a field study by Kaushik etal. (1985) with the pyrethroid permethrin found noacute toxic effects on rotifers at a concentration of5 mg/l and subsequent laboratory studies withRotaria spp. showed a 24 h acute LC50 in excessof 10 000 mg permethrin/l when applied as anemulsifiable concentrate. Another laboratorystudy with the ciliate Tetrahymena sp. exposed tocypermethrin also showed no effects on severalphysiological endpoints at concentrations of 20 mg/l (Larsen, 1997). Apparently, the low sensitivity ofrotifers to pyrethroids is not due to lack of uptakeas studies has shown that the uptake of permethrinin Rotaria spp. is rapid and reaches a concentra-tion of about 3$/104 times that in the water in 2 h.Thus the apparent absence of toxicity may be due

to physiological and/or chemical differences be-tween rotifers and crustaceans (Kaushik et al.,1985). The absence of direct toxic effects onphytoplankton and periphyton was also expected,since laboratory studies with algae exposed topyrethroids have not revealed an EC50 or NOEC-value lower than 70 mg/l (Hill, 1985, 1989).In this study, significant positive relations were

found between the abundance of heterotrophicorganisms and the cypermethrin concentrationand between the chlorophyll-a concentration andthe cypermethrin concentration. These significantrelations were observed after 2!/7 days of cyper-methrin exposure (Fig. 3). The biomass of perifyticalgae (chlorophyll-a ) was also significant positiverelated to the cypermethrin concentration on day11 (Fig. 4). Although there was a significantpositive relation between the abundance of ciliatesand the concentration of cypermethrin during theentire study period, the slope of the regression linegot steeper with time, indicating an increasingdifference between the abundance in the controland the cypermethrin treated enclosures (Fig. 3B).The proliferation of rotifers, protozoans, bacteriaand algae was observed at nominal cypermethrinconcentrations of ]/0.13 mg/l, which correspondedto a concentration giving reductions of more than50% in the abundance of crustaceans (Figs. 1 and2). It is well known that crustaceans, especiallycladocerans, can exert a high grazing pressure onautotrophic and heterotrophic microorganisms(Jurgens, 1994; Christoffersen and Bosselmann,1997) and the importance of the cladocerangrazing for the top!/down control and structureof the algae and microbial community has beendemonstrated in numerous cases (e.g. Christoffer-sen et al., 1993; Jurgens, 1994; Christoffersen,1998). Thus, the increase in rotifers, ciliates,HNF, bacteria and planktonic and perifytic algaein the cypermethrin treated enclosures comparedwith the control enclosures in the present fieldstudy, is most likely due to a reduced top!/downcontrol from the crustaceans. More specificly, thedrastic decrease in the abundance of cladoceransfollowing the application of cypermethrin andhence relaxation of their grazing pressure mostprobably enabled the ciliates, HNF and algae toproliferate within only a few days (Fig. 2). The

Fig. 4. The relation between the log-transformed periphyticchlorophyll-a concentration vs. the log-transformed nominalconcentration of cypermethrin on day 11.

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371 367

cladoceran grazing may have been reduced as aconsequence of both lethal and sublethal effects, asDay et al. (1987) have demonstrated that sublethalconcentrations of 0.01 mg/l and ]/0.05 mg fenva-lerate/l reduces the filtering rates of Ceriopadaph-nia lacustris and Daphnia galeata mendotae ,respectively. Later in the course of the experimentthe bacteria proliferated in the high treatmentenclosures which may have been due to thesubstantial numbers of dead and dying crustaceansand the increased algal biomass which enhancedthe availability of organic substrates (Fig. 2D).Furthermore, on the two last sampling dates aclear increase in the abundance of rotifers wasfound (Fig. 2A), probably due to decreased foodcompetition and mechanical interference withDaphnia (Gilbert, 1988) and reduced predationby cypermethrin sensitive carnivorous crustaceans.The increased rotifer abundance and hence in-creased grazing pressure may thus be the explana-tion behind the decrease in the abundance ofciliates and HNF in the enclosures treated with]/0.13 mg cypermethrin/l on the last samplingdates (Fig. 2B, C), even though their potentialfood concentration in the form of bacteria, in-creased.

In support of our findings, increases in theabundance of rotifers coupled with a reduction inthe abundance of crustaceans has been observed inseveral other field experiments in response topyrethroid insecticides (Kaushik et al., 1985; Dayet al., 1987; Fairchild et al., 1992; Lozano et al.,1992; Webber and Deutsch, 1992). Farmer et al.(1995) exposed pond mesocosms to spray-driftdeposition of 0.7 g a.i. cypermethrin/ha on fouroccasions, each of the applications resulting in anominal exposure concentration of approximately0.07 mg cypermethrin/l (our calculation), andfound an increase in the total number of rotifers,phytoplankton and periphytic algal cells and totalbiomass on some of the sampling occasions.However, no effects on the abundance of zoo-plankton were found in their study, rather theauthors attributed the effects observed on therotifers and algal community to indirect effectsresulting from a decreased abundance of macro-invertebrates. Contrary to the result from thepresent study, Farmer et al. (1995) did not observe

any effect on the concentration of chlorophyll-a orthe community structure of phytoplankton (foreffects on community structure in the presentstudy see Wendt-Rasch et al., 2002). Increasedmortality of aquatic insects and crustacean zoo-plankton and the development of a filamentousalgal mat was found by Crossland et al. (1982)following exposure of experimental ponds tonominal concentrations of cypermethrin of 12.5!/

14.3 mg/l. These changes resemble the alterationsencountered in this experiment; although thepresent experiment indicates that effects may occurat nominal concentrations one hundred timeslower (0.13 mg/l). Based on recorded effects onplankton communities with 12 different insecti-cides in 19 different mesocosm studies it wasconcluded that indirect effects are very variablein both magnitude and direction (Anonymous,2000). This variability may stem from the fact thatthe outcome of indirect effects is highly dependenton the actual species composition of the ecosystemin question. The most prominent indirect effects ofinsecticides on plankton communities include in-creases in phytoplankton and rotifers (Anon-ymous, 2000). However, as indirect effects onprotozoans and bacteria hardly have been investi-gated in mesocosm studies it is yet unknown howoften effects of insecticide exposure propagate tothese components. Our study shows that indirecteffects on these components including rotifers andalgae may occur within only 2!/7 days afterrelaxation of grazing and competition from crus-taceans due to the very short generation times ofthe microbial components.Cypermethrin is seldom recorded in natural

waters possibly due to its rapid disappearancefrom the dissolved phase in the water column.However, pyrethroids with a comparable recom-mended dosage and use range as cypermethrinhave been found in natural waters in detectableconcentrations of 0.66!/0.03 mg esfenvalerate/l(Anonymous, 1999; Wiggers, 1999), 0.11 mg fen-valerate/l and 0.13 mg permethrin/l (Cooper, 1991).Predicted environmental concentrations (PEC) ofcypermethrin in surface waters, assumed to becaused by spray-drift over 0.25 m deep water, havebeen estimated to 0.02!/3 mg cypermethrin/l fol-lowing a single application (Linders et al., 2002).

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371368

Cypermethrin concentrations of this order ofmagnitude may, therefore, be expected to occurin natural waters at least for a short time,following spraying events. Thus, the results ofthe present study verify that expected environ-mental concentrations of pyrethroids is a factorthat may interrupt the crustaceans top!/downcontrol of the microbial community including thealgal community. The insecticide exposure in thepresent study resulted in a shift in the structure ofthe planktonic community, from dominance by thelarge crustaceans to dominance by the smaller, lessgrazer efficient, rotifers and protozoans and alsoin an increase in the abundance of bacteria andalgae. These indirect effects of cypermethrin mayhave further implications at the ecosystem level. Achange in plankton community composition to-wards dominance of microzooplankton will dimin-ish the role of the grazer food web in favour of thedecomposer food web which in turn may influencethe carbon and energy transfer in the aquaticecosystem as the food chain gets longer (Riemannand Christoffersen, 1993). Furthermore, the re-duced grazer control from crustacean zooplanktonmay cause a shift from a macrophyte- to an algae-dominated state of the ecosystem as an increasedbiomass of phytoplankton and periphyton mayreduce the available light to submersed plants andthereby reduce their growth (Sand-Jensen andBorum, 1991; Weisner et al., 1997). Contamina-tion of water bodies with pyrethroids like cyper-methrin may thus add to the adverse effects causedby increased nutrient load and may be a factorcounteracting restoration from a eutrophic state.

Acknowledgements

We thank Anna Arvidson for performing theanalysis of cypermethrin and Erik Jørgensen fortechnical assistance. This work was supported bythe University of Copenhagen, the Øresund Uni-versity, the Swedish Foundation for StrategicEnvironmental Research (MISTRA), the SwedishCouncil for Forestry and Agricultural Research(SJFR) and the Swedish Biodiversity Centre(CBM).

References

Anderson, R.L., 1989. Toxicity of synthetic pyrethroids tofreshwater invertebrates. Environ. Toxicol. Chem. 8, 403!/410.

Anonymous, 1999. Bichelrapport, Rapport fra underudvalgetom Miljø og sundhed. Danish Environmental Pro-tection Agency, Ministry of Environment and Energy,Denmark.

Anonymous, 2000. Guidelines for evaluating mesocosm experi-ments in connection with the approval procedure. A sub-project of Model-based tool for evaluation of exposure andeffects of pesticides in surface water. VKI, Institute for theWater Environment, Division of Ecology, Department ofMonitoring and Information Technology and NationalEnvironmental Research Institute, Department of Lakeand Estuarine Ecology, Denmark.

Bloomquist, J.R., 1996. Ion channels as targets for insecticides.Ann. Rev. Entomol. 41, 163!/190.

Brock, T.C.M., Budde, B.J., 1994. On the choice of structuralparameters and endpoints to indicate responses of fresh-water ecosystems to pesticide stress. In: Hill, I.R., Heim-bach, F., Leeuwangh, P., Matthiessen, P. (Eds.), FreshwaterField Tests for Hazard Assessment of Chemicals. LewisPublisher, Boca Raton, FL, pp. 19!/56.

Christoffersen, K., 1998. Abundance, size, and growth ofheterotrophic nanoflagellates in eutrophic lakes with con-trasting Daphnia and macrophyte densities. In: Jeppesen,E., Søndergaard, M., Christoffersen, K. (Eds.), The Struc-turing Role of Submerged Macrophytes in Lakes. Springer,USA, pp. 274!/282.

Christoffersen, K., Bosselmann, S., 1997. Zooplankton:growth, grazing and interactions with fish. In: Sand-Jensen,K., Pedersen, O. (Eds.), Freshwater Biology, Prioritiesand Development in Danish Research (On the one hun-dredth anniversary of the Danish Freshwater BiologicalLaboratory). GEC Gad, Copenhagen, Denmark, pp. 162!/182.

Christoffersen, K., Riemann, B., Klysner, A., Søndergaard, M.,1993. Potential role of fish predation and natural popula-tions of zooplankton in structuring a plankton communityin eutrophic lake water. Limnol. Oceanogr. 38, 561!/573.

Cooper, C.M., 1991. Insecticide concentrations in ecosystemcomponents of an intensively cultivated watershed inMississippi. J. Freshwater Ecol. 6, 237!/247.

Crossland, N.O., Shires, S.W., Bennett, D., 1982. Aquatictoxicology of cypermethrin. III. Fate and biological effectsof spray drift deposits in fresh water adjacent to agriculturalland. Aquat. Toxicol. 2, 253!/270.

Davies, J.H., 1985. The pyrethroids: an historical introduction.In: Leahey, J.P. (Ed.), The Pyrethroid Insecticides. Taylor &Francis, London and Philadelphia, pp. 1!/31.

Day, K.E., 1989. Acute, chronic and sublethal effects ofsynthetic pyrethroids on freshwater zooplankton. Environ.Toxicol. Chem. 8, 411!/416.

Day, K.E., Kaushik, N.K., 1987a. Short-term exposure ofzooplankton to the synthetic pyrethroid, fenvalerate, and its

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371 369

effects on rates of filtration and assimilation of the alga,

Chlamydomonas reinhardii . Arch. Environ. Contam. Tox-

icol. 16, 423!/432.Day, K.E., Kaushik, N.K., 1987b. An assessment of the chronic

toxicity of the synthetic pyrethroid, fenvalerate, to Daphnia

galeata mendotae , using life tables. Environ. Pollut. 44, 13!/26.

Day, K.E., Kaushik, N.K., Solomon, K.R., 1987. Impact of

fenvalerate on enclosed freshwater planktonic communities

and on in situ rates of filtration of zooplankton. Can. J. Fish

Aquat. Sci. 44, 1714!/1728.Draper, N.R., Smith, H., 1981. Applied Regression Analysis.

Wiley, NY, USA.European Commission, 1995. Risk assessment of existing

substances. Technical guidance document in support of

the commission regulation (EC) no. 1488/94 on risk assess-

ment for existing substances in accordance with council

regulation (EEC) no. 793/93. European Commission, Di-

rectorate-general, Environment, nuclear safety and civil

protection.Fairchild, J.F., La Point, T.W., Zajicek, J.L., Nelson, M.K.,

Dwyer, F.J., Lovely, P.A., 1992. Population-, community-

and ecosystem-level responses of aquatic mesocosms to

pulsed doses of a pyrethroid insecticide. Environ. Toxicol.

Chem. 11, 115!/129.Farmer, D., Hill, I.R., Maund, S.J., 1995. A comparison of the

fate and effects of two pyrethroid insecticides (lamda-

cyhalothrin and cypermethrin) in pond mesocosms. Ecotox-

icology 4, 219!/244.Gilbert, J.J., 1988. Supression of rotifer populations by

Daphnia : a review of the evidence, the mechanisms, and

the effects on zooplankton community structure. Limnol.

Oceanogr. 33, 1286!/1303.Gulati, R.D., 1990. Structural and grazing responses of

zooplankton community to biomanipulation of some Dutch

water bodies. Hydrobiologia 200/201, 99!/118.Haya, K., 1989. Toxicity of pyrethroid insecticides to fish.

Environ. Toxicol. Chem. 8, 381!/391.Hill, I.R., 1985. Effects on non-target organisms in terrestrial

and aquatic environments. In: Leahey, J.P. (Ed.), The

Pyrethroid Insecticides. Taylor & Francis, London and

Philadelphia, pp. 151!/262.Hill, I.R., 1989. Aquatic organisms and pyrethroids. Pestic. Sci.

27, 429!/465.Hill, I.R., Shaw, J.F., Maund, S.J., 1994. Review of aquatic

field tests with pyrethroid insecticides. In: Hill, I.R.,

Heimbach, F., Leeuwangh, P., Matthiessen, P., (Eds.),

Freshwater Field Tests for Hazard Assessment of Chemi-

cals. Lewis Publisher, Boca Raton, FL, pp. 249!/271.Hough, R.A., Fornwall, M.D., Negele, B.J., Thompson, R.L.,

Putt, D.A., 1989. Plant community dynamics in a chain

of lakes: principal factors in the decline of rooted macro-

phytes with eutrophication. Hydrobiologia 173, 199!/217.

Hurlbert, S.H., Mulla, M.S., Willson, H.R., 1971. Effects of an

organophosphorus insecticide on the phytoplankton, zoo-

plankton, and insect populations of fresh-water ponds.

Ecol. Monogr. 42, 269!/299.International Standard Organization, 1996. Water quality*/

Determination of the inhibition of the mobility of Daphnia

magna Straus (Cladocera, Crustacea )-Acute toxicity test,

International Standard Organization (ISO) 6341.Jespersen, A.M., Christoffersen, K., 1987. Measurements of

chlorophyll-a from phytoplankton using ethanol as extrac-

tion solvent. Arch. Hydrobiol. 109, 445!/454.Jurgens, K., 1994. Impact of Daphnia on planktonic microbial

food webs*/a review. Mar. Microb. Food Webs 8, 295!/324.

Kaushik, N.K., Stephenson, G.L., Solomon, K.R., Day, K.E.,

1985. Impact of permethrin on zooplankton communities in

limnocorrals. Can. J. Fish Aquat. Sci. 42, 77!/85.Kjølholt, J., Schou, L., Hald, C., 1991. Økotoksikologisk

vurdering af cypermethrin (ICI). COWIconsult, Danish

Environmental Protection Agency, Ministry of Environ-

ment and Energy, Denmark, Document no. 20754-01.Lampert, W., Fleckner, W., Rai, H., Taylor, B.E., 1986.

Phytoplankton control by grazing zooplankton: a study

on the spring clear-water phase. Limnol. Oceanogr. 31,

478!/490.Larsen, J., 1997. Effects of Cypermethrin, Pirimicarb and

Dimethoate on Tetrahymena . Pesticices Research, 35.

Danish Environmental Protection Agency, Ministry of

Environment and Energy, Denmark.Larson, S.J., Gilliom, R.J., Capel, P.D., 1999. Pesticides in

streams of the United States-initial results from the national

water-quality assessment program. US Geological Survey,

Water-Resource Investigation Report 98-4222.Leahey, J.P., 1985. Metabolism and environmental degrada-

tion. In: Leahey, J.P. (Ed.), The Pyrethroid Insecticides.

Taylor & Francis, London and Philadelphia, pp. 263!/343.Liber, K., Narinder, K.K., Solomon, K.R., Carey, J.H., 1992.

Experimental designs for aquatic mesocosm studies: A

comparison of the ‘ANOVA’ and ‘Regression’ design for

assessing the impact of tetrachlorophenol on zooplankton

populations in limnocorrals. Environ. Toxicol. Chem. 11,

61!/77.Linders, J.B.H.J., Jansma, J.W., Mensink, B.J.W.G., Otter-

mann, K., 2002. Pesticides: benefaction or pandora’s box? A

synopsis of the environmental aspects of 243 pesticides,

National Institute of public health and environmental

protection, Bilthoven, The Netherlands, Report no.

679101014.Lozano, S.J., O’Halloran, S.L., Sargent, K.W., Brazner, J.C.,

1992. Effects of esfenvalerate on aquatic organisms in

littoral enclosures. Environ. Toxicol. Chem. 11, 35!/47.Mian, L.S., Mulla, M.S., 1992. Effects of pyrethroid insecticides

on nontarget invertebrates in aquatic ecosystems. J. Agric.

Entomol. 9, 73!/98.Miller, T.A., Salgado, V.L., 1985. The mode of action of

pyrethroids on insects. In: Leahey, J.P. (Ed.), The Pyre-

throid Insecticides. Taylor & Francis, London and Phila-

delphia, pp. 43!/97.

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371370

Phillips, G.L., Eminson, D., Moss, B., 1978. A mechanism to

account for macrophyte decline in progressively eutrophi-

cated freshwaters. Aquat. Bot. 4, 103!/126.Porter, K.G., Feig, Y.S., 1980. The use of DAPI for identifying

and counting aquatic microflora. Limnol. Oceanogr. 25,

943!/948.Riemann, B., Christoffersen, K., 1993. Microbial trophody-

namics in temperate lakes. Mar. Microb. Food Webs 7, 69!/100.

Sand-Jensen, K., Borum, J., 1991. Interactions among phyto-

plankton, periphyton, and macrophytes in temperate fresh-

waters and estuaries. Aquat. Bot. 41, 137!/175.Sharom, M.S., Solomon, K.R., 1981. Adsorption-desorption,

degradation, and distribution of permethrin in aqueous

systems. J. Agric. Food. Chem. 29, 1122!/1125.Shires, S.W., Bennett, D., 1985. Contamination and effects in

freshwater ditches resulting from an aerial application of

cypermethrin. Ecotoxicol. Environ. Saf. 9, 145!/158.Soderlund, D.M., Bloomquist, J.R., 1989. Neurotoxic actions

of pyrethroid insecticides. Ann. Rev. Entomol. 34, 77!/96.

Ulen, B., Kreuger, J., Sundin, P., 2002. Undersokning avbekampningsmedel i vatten fran jordbruk och samhallen ar2001. Ecohydrologi 4, 1!/30.

Webber, E.C., Deutsch, W.G., 1992. Ecosystem-level testing ofa synthetic pyrethroid insecticide in aquatic mesocosms.Environ. Toxicol. Chem. 11, 87!/105.

Weisner, S.E.B., Strand, J.A., Sandsten, H., 1997. Mechanismsregulating abundance of submerged vegetation in shalloweutrophic lakes. Oecologia 109, 592!/599.

Wendt-Rasch, L., Friberg-Jensen, U., Christoffersen, K., Woin,P., 2002. Direct and indirect effects of the pyrethroidinsecticide cypermethrin on a freshwater plankton commu-nity studied under field conditions. II. Primary Producers,Accepted for publication in Aquatic Toxicology.

Wiggers, L., 1999. Pesticider i vandløb, kilder og søer i ArhusAmt. Teknisk rapport. Arhus Amt, Natur og Miljøkon-toret, Denmark.

Woin, P., 1994. C8 Solid-phase extraction of the pyrethroidinsecticide fenvalerate and the chloroacetanilide herbicidemetazachlor from pond water. Sci. Total Environ. 156, 67!/75.

U. Friberg-Jensen et al. / Aquatic Toxicology 63 (2003) 357!/371 371