single- and two-species tests to study effects of the anthelmintics ivermectin and morantel and the...

316

Environmental Toxicology and Chemistry, Vol. 28, No. 2, pp. 316–323, 2009� 2009 SETAC

Printed in the USA0730-7268/09 $12.00 � .00

SINGLE- AND TWO-SPECIES TESTS TO STUDY EFFECTS OF THE ANTHELMINTICSIVERMECTIN AND MORANTEL AND THE COCCIDIOSTATIC MONENSIN ON

SOIL INVERTEBRATES

JOHN JENSEN,*† XIAOPING DIAO,‡ and ANNE DUUS HANSEN§†Danish National Environmental Research Institute, Aarhus University, P.O. Box 314, Vejlsøvej 25, DK-8600 Silkeborg, Denmark

‡College of Life Science and Agriculture, Hainan University, Haikou 570228, China§Department of Biology, Aarhus University, 8000 Aarhus, Denmark

(Received 10 December 2007; Accepted 30 July 2008)

Abstract—Soil invertebrates in arable land are potentially exposed to veterinary medicines excreted by husbandry. The toxicityof three widely used pharmaceuticals was therefore investigated with the use of common soil invertebrates exposed in the laboratoryin single- or two-species test system. The anthelmintic morantel did not cause significant mortality to either Folsomia fimetaria orEnchytraeus crypticus even at the highest tested concentration of 900 mg kg�1 dry soil. The coccidiostatic monensin affected thereproduction of F. fimetaria and E. crypticus with soil concentrations estimated to cause a 10% effect at values of approximately109 and 71.8 mg kg�1 dry soil, respectively, but caused no mortality to adult. The anthelmintic ivermectin did not affect the survivalof adult Hypoaspis aculeifer. Reproduction of H. aculeifer declined approximately 45% in response to ivermectin exposure of 5mg kg�1 dry soil. Ivermectin was highly toxic to F. fimetaria and affected the survival of adults with soil concentrations estimatedto cause a 50% mortality at values of 5.3 mg kg�1 dry soil in the single-species test system and 0.14 mg kg�1 dry soil in the two-species test system. Reproduction of F. fimetaria was reduced by ivermectin with 10% effective concentration at 0.19 mg kg�1 drysoil in the single-species test system and 0.02 mg kg�1 dry soil in two-species test system. It was shown that species interactionsmay influence the response of test organisms to toxic substances. The data from this study and previously published data showedthat, whereas ivermectin is likely to pose a risk to soil-dwelling invertebrates, adverse effects of morantel and monensin are unlikelyto occur as a result of residue excretion from treated farm animals.

Keywords—Veterinary pharmaceutical Ecotoxicity Springtails Enchytraeids Predatory mites

INTRODUCTION

Veterinary medicinal products (VMPs) are used worldwideto treat diseases, protect animal health, enhance productivity,and promote growth. Anthelmintics and coccidiostatics are twolarge groups of VMPs extensively used in modern farming.Veterinary pharmaceuticals may reach the soil either directlyvia dung on pasture or indirectly via application of manureand slurry as field fertilizer. According to Halling-Sørensen etal. [1] and Boxall et al. [2,3], ivermectin, morantel, and mo-nensin all have high potential for entering the environment.Endoparasiticides such as morantel and ivermectin are usedto treat grazing animals, as well as intensively bred animalskept in stables, whereas coccidiostatics such as monensin areprimarily used as food additives for stabled animals. Manurefrom animals kept in stables is often ploughed into arable soiland may hereafter pose a hazard to soil-living organisms. Fur-thermore, leaching from dung pats may reach the underlyingsoil. Although springtails, enchytraeids, and mites are foundin fresh dung pats only to a limited amount, they may activelyseek dung pats when they are partly degraded but still containincreased levels of the slowly degradable antiparasitics.

Morantel is a tetrahydropyrimidine anthelmintic and is ad-ministered as an endoparasiticide to sheep and cattle to treatgastrointestinal roundworms and tapeworms [4; www.emea.europa.eu/pdfs/vet/mrls/087503en.pdf]. No trans-Europeanconsumption data have been identified; however, the distrib-

* To whom correspondence may be addressed ([email protected]).Published on the Web 9/23/3008.

uted annual quantity of morantel in the United Kingdom wasat least 2,986 kg of active compound in 2000 [5; publications.environment-agency.gov.uk/pdf/SP6-012-8-TR-e-p.pdf]. Mor-antel acts as an agonist at nicotic acetylcholine receptors onthe muscle cells of nematodes. The binding of morantel to thereceptor results in depolarization of the membranes in musclecells, leading to spastic paralysis of the worms [6].

Ivermectin belongs to the macrocyclic lactone group ofanthelmintics, with broad-spectrum activity against endo- andectoparasites such as gastrointestinal roundworms, lungworms,cattle grubs, mites, and lice [7]. Ivermectin is used to treatfarm animals, as well as pets. Ivermectin was the most soldendoparasiticide in the United Kingdom in 2000, with an an-nual amount of at least 3,995 kg of active compound [5]. Thisnumber, however, does not account for other routes of appli-cation, for instance, treatment of ectoparasites. Ivermectin isbelieved to act by binding with glutamate-gated chloride chan-nels in invertebrates’ (such as nematodes and arthropods)nerve and muscle cells. This, in turn, leads to increased chlo-ride permeability, hyperpolarization of the cell membrane, andultimately paralysis and death of the organism. Ivermectin mayalso interfere with the �-aminobutyric acid (GABA) receptor–chloride channel [8,9].

Monensin is an ionophore coccidiostatic and exhibits bothantibacterial and anticoccidial activities. However, monensinis primarily used to control the parasitic disease coccidiosiscaused by Eimeria species in poultry. A secondary use is inthe improvement of feed efficiency and weight gain in cattle[10,11]. Monensin acts as an antiporter facilitating the move-ment of monovalent cations (sodium and potassium) across

Effects of veterinary pharmaceuticals to soil invertebrates Environ. Toxicol. Chem. 28, 2009 317

the membranes of both prokaryotic and eukaryotic cells inexchange for a proton. This disrupts the ion gradient, causingenergy deficiency and cell death [10]. In the European Union,some coccidiostatics (including monensin) are permitted asfeed additives for poultry production, and they are extensivelyused in Denmark, where the annual consumption exceeded13,000 kg of active compound in 2004 (www.danmap.org).

Morantel, ivermectin, and monensin are mainly eliminatedvia feces [12–15]. A substantial proportion of ivermectin andmorantel is excreted as parent drug, whereas a larger propor-tion of monensin may be metabolized depending on the treatedanimal [15–17]. Furthermore, as both morantel and ivermectindegrade relatively slowly in dung [18–21], their use may raiseconcern for the terrestrial environment. Slowly degrading sub-stances have the potential of reaching surface water by runoff.Monensin, for example, has been detected in surface waterreceiving runoff from agriculture land in Canada [22].

The aim of the present study was to investigate toxic effectsof the widely used veterinary pharmaceuticals—ivermectin,morantel, and monensin—on the soil invertebrate fauna, usingthe springtail (Folsomia fimetaria), the enchytraeid (Enchy-traeus crypticus), and the predatory mite (Hypoaspis aculei-fer). Although intensively studied, ivermectin was selected asan excellent model substance to evaluate the effects of speciesinteractions and chemical stress, whereas the toxicity of mor-antel was tested to elucidate the environmental properties ofan alternative antiparasitic compound. Monensin was selectedas test compound as a representative for a widely used classof veterinary medicines (i.e., the coccidiostatics).

Springtails, enchytraeids, and mites are some of the mostabundant and widespread soil-dwelling invertebrates [23].They are ecologically important in soils, contributing to thedecomposition of organic materials, nutrient release, and for-mation of soil structure [24–26]. The springtail F. fimetariaand the predatory mite H. aculeifer are both common inhab-itants in agricultural soil in temperate regions [27,28]. E. cryp-ticus and other Enchytraeus species are typically associatedto the litter layer of soils or organic matrices like compost anddung, where they can be found in high numbers [24,29]. Theyare widely distributed, and more than 100 species have beenidentified in Europe alone.

In the present study, the effects of monensin and morantelwere evaluated on springtails and enchytraeids. The effects ofivermectin were assessed with springtails in a single-speciestest and with springtails and mites in a two-species predator–prey test.

MATERIALS AND METHODS

Test species

All test animals were supplied from stock cultures at theDanish National Environmental Research Institute (NERI), andkept at 20 � 1�C with a 12:12-h light:dark photoperiod.

Folsomia fimataria (Collembola: Isotomidae) is a eue-daphic, eyeless, nonpigmented, sexually reproducing spring-tail. The animals were bred on Paris-charcoal plaster in Petridishes and fed dried baker’s yeast. Synchronous animals wereobtained in accordance with Wiles and Krogh [28]. Animalsbetween 23 to 26 d old were used in the single-species ex-periment.

E. crypticus (Oligochaeta: Enchytraeidae) is a white her-maphroditic potworm. The worms were bred on a Bacto agarsubstrate in Petri dishes, and they were fed oatmeal. Only

sexually mature animals with visible clitellum and of approx-imately the same size were used in the experiment.

H. aculeifer (Gamasida: Laelapidae) is a hemiedaphic oreuedaphic polyphagous predatory mite. It has an arrhenoto-kous mode of reproduction, whereby unfertilized females pro-duce male offspring. Synchronism was obtained in accordancewith Krogh and Axelsen [30]. Juvenile mites were fed juvenilespringtails (Folsomia candida) every second day. Mites of 16to 19 d old were used in the test.

Test compound

Monensin sodium salt (Chemical Abstracts Service [CAS]22373-78-0) and morantel citrate salt (CAS 69525-81-1) wereobtained from Sigma-Aldrich (Brøndby, Denmark). Ivermectin(CAS 70288-86-7) was obtained from MP Biomedicals (Irvine,CA, USA). The purity of the chemicals was 90 to 95, 99.5,and 95% for monensin, morantel, and ivermectin, respectively.

Test soil

The test soil was sampled at an organic agriculture field(A horizon) at Sjællands Odde, Denmark. The test soil rep-resents a typical Danish agricultural soil with no recent historyof pesticide use. To exclude intrinsic soil animals, the soil wasdried at 80�C for 24 h and thereafter sieved through a 2-mmmesh to remove larger particles. The test soil was a sandy clayloam with a pH of 7.0, and the soil had a particle size distri-bution of 67% sand, 12% silt, and 21% clay. The total carboncontent of the soil was 2.22%.

Test procedure

Generally, the tests were conducted according to internalstandard operation procedures at the NERI. These standardoperation procedures are based on international (Organisationfor Economic Co-operation and Development, Paris, France;International Organization for Standardization, Geneva, Swit-zerland) standard test guidelines when available. However,some tests were slightly modified versions of the guidelinesto optimize them for practicality, selected species, and chosenendpoints.

Springtail tests. The toxicity tests were performed accordingto the method proposed by Wiles and Krogh [28], which is ingeneral agreement with International Organization for Stan-dardization guideline 11267 (www.iso.org/iso/iso�catalogue.htm). Test substances were dissolved in acetone (ivermectinand monensin) or demineralized water (morantel). Series offour, five, or six test concentrations and controls each withfour replicates were used. For ivermectin and monensin, onlyacetone was added to the control soil. Long-time experiencein the laboratory of NERI has demonstrated that the presenttest procedure allows all acetone to evaporate before animalsare added; therefore, a water control is considered unnecessary.Test concentrations were 0.1, 0.2, 0.4, 1.2, 2.5, and 5 mg kg�1

dry soil for ivermectin, 60, 125, 250, 500, and 800 mg kg�1

dry soil for monensin, and 200, 400, 600, and 900 mg kg�1

dry soil for morantel. The ivermectin concentrations were se-lected on the basis of previous studies with soil invertebrates(Table 1), whereas the other test concentrations were selectedon the basis of short-range finding tests.

In tests where acetone was used as solvent, the acetone wasallowed to evaporate overnight in a fume cupboard and thenext day demineralized water was added to the soil. Studiesof the soil structure and water film under microscope hadidentified the amount of water needed to create optimal con-

318 Environ. Toxicol. Chem. 28, 2009 J. Jensen et al.

Table 1. Estimated concentration causing 10 and 50% reductions in reproduction (EC10 and EC50), and 50% mortality (LC50) of adult springtails(Folsomia fimetaria), enchytraeids (Enchytraeus crypticus), and mites (Hypoaspis aculeifer)a

EC10 (mg kg�1 dry soil) EC50 (mg kg�1 dry soil) LC50 (mg kg�1 dry soil) NOECb (mg kg�1 dry soil)

IvermectinSpringtails 0.19 [�0.01; 0.38] 0.93 [0.52; 1.35] 5.3c [4.46; 6.63] 0.4Enchytraeids 14 [6; 22]d 36 [25; 48]d �300d 3d

Two-species testSpringtails (prey) 0.02 [�0.09; 0.13] 0.11 [�0.10; 0.32] 0.14 [0.01; 0.27] 0.2Predatory mites 0.04 [0.03; 0.08] �5 �5 �5

MonensinSpringtails 109 [�80; 298] 590.7 [254; 927] �800 250Enchytraeids 71.8 [�65; 209] 356 [95; 617] �800 500

MorantelSpringtailse �900 �900 �900 �900Enchytraeidse �900 �900 �900 �900

a The 95% confidence limits are presented in square brackets.b NOEC � no-observed-effect concentration.c The LC50 value is obtained by extrapolation beyond the maximum test concentration; it was possible to estimate an LC45 value of 4.8 mg

kg�1.d Data published by Jensen et al. [37] using another Danish agricultural soil as test medium.e No clear dose–response relationship could be established.

ditions for the springtails. Thus, 5 ml of solution were mixedinto the 25-g dry soil equivalent to an initial water content of20% on weight basis, which corresponds to approximately 50%of the waterholding capacity of that specific soil type.

The soil was transferred to the test cylinders (5.5 cm high,6 cm diameter, with a 1-mm mesh in the bottom), and theanimals (10 females and 10 males) were added to each rep-licate. In tests where the compound was dissolved in water,animals were added the same day as the soil was mixed. Food(yeast) was added to the test cylinders before they were closedat the top and bottom with lids, and the cylinders were in-cubated for 21 d at 20 � 1�C with a 12:12-h light:dark pho-toperiod.

After two weeks, the samples were weighed and water wasadded to compensate for water loss by evaporation. At thistime, the animals were fed 15 g of baker’s yeast. After in-cubation, the animals were extracted over 3 d in a MacFayden-type temperature gradient extractor [31]. The temperature inthe upper compartment of the extractor was increased stepwisefrom 25 to 40�C (5�C every 12 h), and in the lower compart-ment the temperature was kept constant at 5�C. Animals werecollected and stored at 5�C until they were counted. The an-imals were counted using a digital image processing systemaccording to Krogh et al. [32].

Enchytraeid tests. The tests were generally performed inaccordance with International Organization for Standardiza-tion guideline 16387 (www.iso.org/iso/iso�catalogue.htm) us-ing the species E. crypticus instead of the species E. albidusoriginally included in the guideline. Five test concentrationsand a control were used with four replicates of each concen-tration. Test concentrations were 60, 125, 250, 500, and 800mg kg�1 dry soil for monensin and 100, 200, 400, 600, and900 mg kg�1 dry soil for morantel. The compounds were dis-solved in demineralized water or acetone, and 5 ml of solutionwas mixed homogeneously into 20 g of dry soil correspondingto an initial water content of 25%. In tests where acetone wasused, the acetone was allowed to evaporate overnight in a fumecupboard before 5 ml of demineralized water was added. Thiscorresponds approximately to 60% of the waterholding ca-pacity of that specific soil type. The wet soil was then trans-

ferred to test cylinders (5 cm high, 3.5 cm inner diameter),and 10 animals were added and were feed dried oatmeal. Thetest containers were closed with perforated lids. The enchy-traeids were exposed for 21 d at 20 � 1�C with a 12:12-hlight:dark photoperiod.

After 14 d, the animals were fed oatmeal and water wasadded to compensate for any water lost by evaporation. Afterincubation, animals were extracted by dividing each sampleinto four containers and water was added. The containers wereclosed and shaken gently for approximately 5 s; then they wereleft for sedimentation for 24 h at 5�C. The next day, containerswere investigated under a microscope, and the animals foundon the surface of the sedimented soil particles were transferredto Petri dishes. Each of the four subsamples was counted man-ually under a microscope, and the numbers were summed upfor each replicate.

Two-species test (Predatory mite and springtail). The testwas performed according to the method of Krogh and Axelsen[30], using F. fimetaria as prey and H. aculeifer as predator.Test concentrations were 0.2, 0.6, 2.5, and 5 mg kg�1 dry soilfor ivermectin with four replicates and a control. In brief, thetest was carried out in accordance with the springtail test re-garding soil preparation, water content, test container, incu-bation conditions, etc., except that 60 g of wet soil (50 g ofdry soil and 10 ml of water, corresponding to a waterholdingcapacity of 50%) was used instead of 30 g of wet soil. Toavoid initial excessive predation, the prey, 100 synchronizedF. fimetaria (16–19 d old), were added 1 h before the additionof the predator, 10 female and 5 male, H. aculeifer. After 21d, the animals were extracted from the soil. Extraction lasted3 d and started at 25�C with temperature increments of 5�Cfor 12 h. Extracted animals were collected and frozen at�20�C, and afterward they were manually counted.

Statistical analysis

Concentrations causing 10 and 50% reduction in repro-duction (EC10 and EC50) were estimated by the PROC NLINprocedure [33; www.sas.com]. The following regression modelfor the EC10 to EC50 estimation was used: c � e�rd � p,where c is a scale factor, r determines the decay rate, and d


Fig. 1. Dose–response curves for effects of ivermectin on the springtailFolsomia fimetaria in the single-species test. (a) Effects on repro-duction; (b) effects on adult survival. Vertical error bars indicatestandard error of mean.

Fig. 2. Dose–response curves for effects on the two-species test ivermectin exposure. (a) Effects on reproduction of the springtail Folsomiafimetaria; (b) effects on survival of adult springtails; (c) effects on reproduction of the mite Hypoaspis aculeifer; (d) effects on survival of adultfemale mites. Vertical error bars indicate standard error of mean.

is the soil concentration. It was not possible to estimate EC10for reproduction of mites with the PROC NLIN proceduresince the dose–response curve did not fit the sigmoid curve.Instead, EC10 was estimated with a linear interpolation usingthe inhibition concentration approach [34]. Estimation of theconcentration causing 50% survival (LC50) was performed byuse of probit analysis [33]. It was not possible to estimateLC50 of survival data of F. fimetaria (two-species) tests withthe PROBIT analysis. Instead, LC50 of the binomially dis-

tributed mortality data was estimated using a PROC NLMIX-ED procedure [33]. Data were checked for normal distributionwith the Kolmogorov-Smirnov test before the no-observed-effect concentration (i.e., the highest tested concentration withno significant effects) was estimated with analysis of varianceand Dunnett’s tests [33]. A significance level of p � 0.05 wasused.

RESULTS

Mean adult survival in the controls was at least 80% forE. crypticus and F. fimeratia and more than 90% for H. acu-leifer (female). The mean number of juveniles in the replicatedcontrols was more than 360 for F. fimeratia (single-speciestest), 380 for E. crypticus, and 38 for H. aculeifer. Test resultswere therefore considered valid according to internationalguidelines and internal standard operation procedures at NERI.The addition of springtails failed in one replication of theivermectin treatment (0.4 mg kg�1 dry soil) in the springtailbioassay. Data were therefore excluded from analyses as anoutlier, because no springtails were recovered. In the followingsection, each of the tested pharmaceuticals is treated in moredetail.

Ivermectin

In the single-species test, ivermectin affected the repro-duction of springtails (F. fimetaria) at lower concentration thanit affected adult survival (Fig. 1). Reproduction of springtailsin the single-species test system was affected with EC10 andEC50 values at 0.19 and 0.93 mg kg�1 dry soil (Table 1).Ivermectin affected survival of adult F. fimetaria with an LC45value at 4.8 mg kg�1 dry soil. By extrapolating the dose–response curve beyond the test concentrations, it was possibleto determine an LC50 of approximately 5.3 mg kg�1 dry soil(Table 1).

In the two-species test system, reproduction and survivalof adults were approximately equally sensitive endpoints forthe prey F. fimetaria (Fig. 2a and b). However, the presenceof the predatory mite H. aculeifer, in combination with iver-mectin exposure, led to decreased reproduction and survival


Fig. 3. Dose–response curves for effects of monensin on Folsomiafimetaria. (a) Effects on reproduction; (b) effects on survival of adults.Vertical error bars indicate standard error of mean.

Fig. 4. Dose–response curves for effects of monensin on the enchy-traeid Enchytraeus crypticus. (a) Effects on reproduction; (b) effectson survival of adults. Vertical error bars indicate standard error ofmean.

of adult F. fimetaria (prey) compared with the single-speciestest (Figs. 1a and b vs 2a and b; see also Table 1). The numberof juvenile F. fimetaria was significantly reduced in all treat-ments compared to control samples (Dunnett’s test) (i.e., low-est-observed-effect concentration is 0.2 mg kg�1 dry soil), andno juveniles were observed at ivermectin concentrations of 2.5and 5 mg kg�1 dry soil (Fig. 2a). Reproduction of the prey F.fimetaria was affected with EC10 and EC50 values at 0.02and 0.11 mg kg�1 dry soil, respectively (Table 1). The pre-dation pressure on adult F. fimetaria was very high since thenumber of animals was reduced by more than approximately90% in the controls after 21 d. Survival of adult F. fimetariawas affected with an LC50 at 0.14 mg kg�1 dry soil (Table1), which is significantly lower than the LC50 of approximately5.0 mg kg�1 dry soil observed in the single-species test.

No clear dose–response relationship was evident for thereproduction of the predatory mite H. aculeifer. Reproductionof the predator H. aculeifer indicated a decline ( 45%) atall ivermectin concentrations (Fig. 2c), but no significant dif-ferences were found between the control and the tested con-centrations (analysis of variance, p � 0.053). Survival of adultH. aculeifer was unaffected, even at the highest ivermectinconcentration of 5 mg kg�1 dry soil (Fig. 2d).

Morantel

No clear dose–response relationships were found for mor-antel. Therefore, neither EC10 nor EC50 values could be es-tablished within the test concentrations used in this experiment.Figures of dose–response experiments are therefore not pre-sented. Morantel had, nevertheless, a small (20% decline forF. fimetaria and 35% decline for E. crypticus) but not sig-nificant effect on reproduction.

Monensin

The numbers of juvenile and adult F. fimetaria after mo-nensin exposure are depicted in Figure 3. Reproduction of F.fimetaria was a more sensitive endpoint than survival, andEC10 and EC50 values was estimated to be approximately 109and 591 mg kg�1 dry soil, respectively (Table 1). Survival ofadult F. fimetaria was not affected by monensin at the con-centrations tested (Fig. 3b).

Dose–response curves for reproduction and survival of theenchytraeids E. crypticus are given in Figure 4. The EC10 andEC50 were estimated to be approximately 72 and 356 mg kg�1

dry soil, respectively. Large variations were observed in re-production output; consequently, large confidence intervals ofthe effective concentrations were found (Table 1). Monensinwas not toxic to adult E. crypticus at the concentrations tested(Fig. 4b).

DISCUSSION

The potential for veterinary pharmaceuticals to adverselyaffect nontarget organisms has raised concern. Effects of an-tiparasitics on dung insects (Diptera and Coleoptera) have beenthoroughly investigated through the last two decades [35,36],whereas only little is known about the ecotoxicological effectsof pharmaceuticals on soil-dwelling invertebrates [37]. Reportson the effects in more complex test systems are hardly avail-able for any substances, including pharmaceuticals [38,39].

Anthelmintics

The present study demonstrates that the presence of thepredatory mite H. aculeifer increases the effects of ivermectinto the springtail F. fimetaria compared to the traditional single-species test. The estimated EC10, EC50, and LC50 for iver-


mectin in the single-species test (F. fimetaria) in this study(0.19, 0.93, and 5.3 mg kg�1 dry wt) are in agreement withresults reported by Jensen et al. [37]. Using another agriculturaltest soil, these authors found EC10, EC50, and LC50 valuesof 0.26, 1.7, and 8.4 mg kg�1 dry weight, respectively, for F.fimetaria, which were slightly higher than the results presentedin this study. The small differences can presumably be ex-plained by different soils types and soil properties.

The EC10, EC50, and LC50 values were reduced by factorsof approximately 10, 9, and 38, respectively, by the presenceof the predatory mite H. aculeifer in the two-species test sys-tem (Table 1). This finding implies that the species interactionmay have significant synergistic effects with the responses ofsoil invertebrates to chemical stress. Thus, the traditional sin-gle-species test with F. fimetaria may underestimate the effectof toxic compounds compared to the more ecologically rele-vant two-species test. However, the reproducibility of resultsin the two-species test may be less robust [30]. Gaining in-creased ecological realism in ecotoxicological test may oftenresult in reduced reproducibility [38]. It is not clear to whatextent the observed increase in toxicity with the addition of apredator may be attributable to age differences in the collem-bolans used in the two tests. Animals used in the single-speciestest were 23 to 26 d old when added, whereas animals in thetwo-species test were 16 to 19 d old. Although the animalsare sexually mature at this age, they may be more susceptibleto toxicants. Alternatively, the increased effects on the preyin the two-species test could be a result of the extra energycosts of predator avoidance [39] or, probably most likely, themode of action, paralysis, and immobility may result in lowerability of the prey to escape the predator, thereby contributingto the lower effective concentrations of F. fimetaria.

Despite the spectrum activity of ivermectin against para-sites, adult H. aculeifer in the two-species test were almostunaffected by ivermectin at the tested concentrations. How-ever, reproduction of the predatory mite H. aculeifer seemedto be affected, although not significantly, in the presence ofivermectin, either directly by ivermectin exposure or indirectlyby depletion of prey. A previous study of two-species toxicitywith mites (H. aculeifer) and springtails (F. fimetaria) indi-cated that pesticide (dimethoate) exposure indirectly affectedthe reproduction of mites due to a decrease in available prey[40]. Adult H. aculeifer can, however, tolerate starvation forseveral months [41], which can explain that no indirect foodchain effect of ivermectin was observed for adult mites duringthis three-week study. Further investigations are necessary toimprove the interpretation of the predatory–prey relationshipand to validate the responses and reproducibility of the test.

Toxicity of ivermectin to soil-dwelling organisms other thanspringtails is covered to some extent in the literature. Jensenet al. [37] reported the effect of ivermectin on reproductionof the enchytraeid E. crypticus (Table 1). For the earthwormEisenia fetida, EC50 values at 15 and 7 mg kg�1 dry weightand an LC50 value at 325 mg kg�1 dry weight have been found[42,43]. Svendsen et al. [44] concluded that Lumbricus ter-restris was unaffected by ivermectin excreted in cattle dung.Springtails are clearly more sensitive to ivermectin than oli-gochaetes (enchytraeids and earthworms), although the dif-ference in sensitivity may partly be due to different experi-mental conditions and ivermectin formulations. The same trendin sensitivity (springtails � oligochaetes) was found with an-other avermectin (i.e., abamectin) by Diao et al. [45] usingthe same soil type as the present study. However, abamectin

was more toxic to F. fimetaria, with EC10, EC50, and LC50values of 0.05, 0.33, and 0.81 mg kg�1 dry soil, respectively,than was ivermectin in the present study. Abamectin and iver-mectin are closely related, and abamectin differs from iver-mectin by only a single double-bond between the carbons atposition C22 to C23 [7].

Reports of effects of ivermectin to beneficial dung-inhab-iting flies and beetles are in the same range as the toxic levelsfor F. fimetaria [46–50].

Various concentrations of ivermectin have been detected indung, reflecting the different formulations, doses, and routesof application. Concentrations of ivermectin in cattle dung mayreach up to 9 mg kg�1 dry weight for treatment with pour-onapplication [46]. However, concentrations of ivermectin indung after injection do not typically exceed 4 mg kg�1 dryweight [20,46,47]. Ivermectin binds strongly to soil and de-grades relatively slowly, with degradation half-lives of morethan 200 d [20,21,39,51]. Jensen et al. [37] estimated the pre-dicted environmental concentrations to be 1.5 mg kg�1 dry soilin Denmark under worst-case assumptions. It can be concludedthat the levels of ivermectin found in dung pats and soil areof the same order of magnitude as the effect levels found inthe present study and therefore that ivermectin poses a risk tosoil-dwelling insects, at least at an individual level. Field stud-ies are needed to evaluate risk at population and ecosystemlevels.

The present study shows that the antiparasitic agent mor-antel was not toxic to springtails and enchytraeids (EC10 �900 mg kg�1 dry soil). This is in agreement with reports fromMcKellar et al. [18], who observed that the tartrate salt ofmorantel had no effect on the development of the yellow dungfly at a high but nonspecified concentration. Concentrations ofmorantel as high as 96 mg kg�1 dry weight has been found incattle dung after oral administration of 10 mg kg�1 body weight[18]. However, since no effects were observed at 900 mg kg�1

dry soil in the present study, is it therefore unlikely that mor-antel will affect soil fauna in arable land. The lack of toxicityof morantel toward soil-living invertebrates implies that mor-antel, when possible from a veterinarian and pharmaceuticalpoint of view, should be used as an alternative to ivermectin,although morantel does not possess the same spectrum of an-tiparasitic activity as ivermectin and therefore can replace iver-mectin only to a limited degree. Suarez et al. [52] studied theeffects of two other antiparasitics (i.e., moxidectin and dora-mectin) on dung-dwelling species. The number of dung-col-onizing species like Coleoptera, Diptera, microarthropods(e.g., collembolans), and nematodes was negatively affectedby the drugs, with doramectin as the most potent of the two.The measured concentrations causing effects were of the sameorder of magnitude presented for ivermectin in this study.Moxidectin and doramectin are, just like ivermectin, potentbroad-spectrum endectocides of the macrocyclic lactone (mac-rolide) class with the same mode of action (i.e., increasedmembrane permeability to chloride ions). It is evident fromthe available information that this mode of action generallyaffects a target widely common throughout various genera ofinvertebrates.

Coccidiostatics

In spite of the extensive use of coccidiostatics and theirpotential presence in the environment, there is little publishedinformation concerning residues in manure and soil and theireffects on terrestrial invertebrates living in these environments.


Median lethal concentration values for monensin of 112.1and 264 mg kg�1 dry weight (14 d) for the earthworm E. fetidahave been reported [53]. Boxall et al. [3] cited the InternationalCooperation on Harmonisation of Technical Requirements forRegistration of Veterinary Products (VICH) for an no-ob-served-effect concentration value for monensin of 10 mg kg�1

dry weight for earthworms, which is lower than the valuesfound for F. fimetaria (250 mg kg�1 dry soil) and E. crypticus(500 mg kg�1 dry soil) in this study. However, as the datafrom the technical requirements are nonpublished data fromthe dossier, it cannot be verified or evaluated further. The EC50for reproduction of E. crypticus and F. fimetaria in the presentstudy was 356 and 590.7 mg kg�1 dry soil, respectively. Al-though the effective concentrations values are of the sameorder of magnitude, E. crypticus was more sensitive to mo-nensin than F. fimetaria, which is also evident from the dose–response curves (Figs. 3 and 4). The opposite order of sen-sitivity was found when comparing no-observed-effect con-centration values (Table 1), due to the larger variation in theenchytraeid data. Herald et al. [54] found that administeringmonensin to cattle significantly reduced the numbers of theface fly and the horn fly in dung pats.

Donoho [15] reported concentrations of monensin in cattledung of up to 4.7 mg kg�1 dry weight. Monensin degradesrelative quickly in manure under natural weathering conditionsand in field soil but is persistent in dung under anaerobe con-dition [15]. Carlson and Mabury [55] reported half-lives ofmonensin applied to field soil of 3.3 and 3.8 d with and withoutmanure, respectively. No data on measured concentration ofmonensin in soil could be found. However, the European FoodSafety Authority estimated the predicted environmental con-centrations in soil from the use of monensin in chickens to bebetween 0.59 and 1.12 mg kg�1 dry soil [56]. The effect levelsfound in the presented study (EC10 values of 71.8 and 109mg kg�1 dry soil) are therefore significantly higher than theexposure levels that can be predicted in soil and even exceedthe levels observed in dung by a factor of approximately 15.Monensin is therefore not likely to affect springtails and en-chytraeids.

CONCLUSION

In conclusion, the present study shows that the springtailF. fimetaria was more affected toward toxic exposure fromivermectin in the two-species test system containing the pred-atory mite H. aculeifer than in the single-species test. Thissuggests that, although the predatory–prey interactions needbetter understanding, it is relevant to involve species inter-actions in higher-tier environmental risk assessment.

It cannot be eliminated that ivermectin may pose a risk tosoil-dwelling invertebrates like F. fimetaria. No effects wereobserved on soil-dwelling invertebrates from morantel even athigh concentrations. Monensin caused a reduction in repro-duction of the springtail F. fimetaria and the enchytraeid E.crypticus. However, by comparison with the few existing dataconcerning concentrations of monensin in dung and soil, it isunlikely that monensin poses a risk to soil-dwelling inverte-brates under realistic exposure scenarios.

Acknowledgement—Financial support from the European Union pro-ject ERAPharm (contract SSPI-CT-2003-511135 under FP6) is ac-knowledged. We also thank Mark Bayley of Arhus University forcritical comments. Thanks to Paul Henning Krogh of NERI for sta-tistical support and to the technicians from NERI, Zdenek Gavor, Elin

Jørgensen, and Kasten T. Andersen, for their assistance in the labo-ratory.

REFERENCES

1. Halling-Sørensen B, Nielsen SN, Lanzky F, Ingerslev F, LutzhøftHC, Jørgensen SE. 1998. Occurrence, fate and effects of phar-maceutical substances in the environment: A review. Chemo-sphere 36:357–393.

2. Boxall ABA, Fogg LA, Kay P, Blackwell PA, Pemberton EJ,Croxford A. 2003. Prioritisation of veterinary medicines in theUK environment. Toxicol Lett 142:207–218.

3. Boxall ABA, Fogg LA, Blackwell PA, Kay P, Pemberton EJ,Croxford A. 2004. Veterinary medicines in the environment. RevEnviron Contam Toxicol 180:1–91.

4. European Medicines Agency. 2004. Morantel. EMEA/MRL/875/03-Final. Summery Report 2. London, UK.

5. Boxall ABA, Fogg LA, Blackwell PA, Kay P, Pemberton EJ.2002. Review of veterinary medicines in the environment. P6-012/8/TR. R&D Technical Report. Environment Agency, Al-mondsbury, Bristol, UK.

6. Martin RJ. 1997. Modes of action of anthelmintic drugs. Vet J154:11–34.

7. Shoop WL, Mrozik H, Fisher MH. 1995. Structure and activityof avermectins and milbemycins in animal health. Vet Parasitol59:139–156.

8. Dourmishev AL, Dourmishev LA, Schwartz RA. 2005. Ivermec-tin: Pharmacology and application in dermatology. Int J Dermatol44:981–988.

9. Wolstenholme AJ, Rogers AT. 2005. Glutamate-gated chloridechannels and the mode of action of the avermectin/milbemycinanthelmintics. Parasitology 131(Suppl):S85–S95.

10. McGuffey RK, Richardson LF, Wilkinson JID. 2001. Ionophoresfor dairy cattle: Current status and future outlook. J Dairy Sci84(E-Suppl):194–203.

11. Butaye P, Devriese LA, Haesebrouck F. 2003. Antimicrobialgrowth promoters used in animal feed: Effects of less-well-knownantibiotics on gram-positive bacteria. Clin Microbiol Rev 16:175–188.

12. Davison KL. 1998. Monensin absorption and metabolism incalves and chickens. J Agric Food Chem 32:1273–1277.

13. Lanusse CE, Gascon LH, Ranjan S, Prichard RK. 1992. Morantelrate release from a long-acting intraruminal device in cattle: Phar-macokinetics and gastrointestinal distribution. J Vet PharmacolTher 15:117–123.

14. Chiu SHL, Green ML, Baylis FP, Eline D, Rosegay A, Meri-wethwe H, Jacob TA. 1990. Absorption, tissue distribution, andexcretion of tritium-labelled ivermectin in cattle, cheep, and rat.J Agric Food Chem 38:2072–2078.

15. Donoho AL. 1984. Biochemical studies of the fate of monensinin animals and in the environment. J Anim Sci 58:1528–1539.

16. Lifschitz A, Virkel G, Sallovitz J, Sutra JF, Galtier P, AlvinerieM, Lanusse C. 2000. Comparative distribution of ivermectin anddoramectin to parasite location tissues in cattle. Vet Parasitol 87:327–338.

17. Lumaret JP, Errouissi F. 2002. Use of anthelmintics in herbivoresand evaluation of risks for the non target fauna of pastures. VetRes 33:547–562.

18. McKellar QA, Scott EW, Baxter P, Anderson LA, Bairden K.1993. Pharmacodynamics, pharmacokinetics and faecal persis-tence of morantel in cattle and goats. J Vet Pharmacol Ther 16:87–92.

19. McKellar QA. 1997. Ecotoxicology and residues of anthelminticcompounds. Vet Parasitol 72:413–435.

20. Sommer C, Steffansen B. 1993. Changes with time after treatmentin the concentration of ivermectin in fresh cow dung and in cowpats aged in the field. Vet Parasitol 48:67–73.

21. Suarez VH, Lifschitz AL, Sallovitz JM, Lanusse CE. 2003. Ef-fects of ivermectin and doramectin faecal residues on the inver-tebrate colonization of cattle dung. J Appl Entomol 127:481–488.

22. Hao CY, Lissemore L, Nguyen B, Kleywegt S, Yang P, Keith S.2005. Determination of pharmaceuticals in environmental watersby liquid chromatography/electrospray ionization/tandem massspectrometry. Anal Bioanal Chem 384:505–513.

23. Petersen H, Luxton M. 1982. A comparative analysis of soil faunapopulations and their role in decomposition processes. Oikos 39:287–388.


24. Didden WAM. 1993. Ecology of terrestrial Enchytraeidae. Pe-dobiologia 37:2–29.

25. Løkke H, van Gestel CAM. 1998. Handbook of Soil InvertebrateToxicity Tests. John Wiley, Chichester, UK.

26. Rusek J. 1998. Biodiversity of Collembola and their functionalrole in the ecosystem. Biodivers Conserv 7:1207–1219.

27. Usher MB, Davis PR. 1983. The biology of Hypoaspis aculeifer(Canestrini) (Mesodtigmata): Is there a tendency towards socialbehaviour? Acarologia 24:243–253.

28. Wiles JA, Krogh PH. 1998. Testing with the collembolans I.viridis, F. candida and F. fimetaria. In Løkke H, van GestelCAM, eds, Handbook of Soil Invertebrate Toxicity Tests. JohnWiley, Chichester, UK, pp 131–156.

29. Jansch S, Rombke J, Didden W. 2005. The use of enchytraids inecological classification and assessment concepts. Ecotoxicol En-viron Saf 62:266–277.

30. Krogh PH, Axelsen JA. 1998. Test on the predatory mite Hy-poaspis aculeifer preying on the collembolan Folsomia fimetaria.In Løkke H, van Gestel CAM, eds, Handbook of Soil InvertebrateToxicity Tests. John Wiley, Chichester, UK, pp 131–156.

31. Petersen H. 1978. Some properties of the two high gradient ex-tractors for soil mocroarthropods. Nat Jutl 20:95–121.

32. Krogh PH, Johansen K, Holmstrup M. 1998. Automatic countingof collembolans for laboratory experiments. Appl Soil Ecol 7:201–205.

33. SAS Institute. 1999. SAS/STAT� User’s Guide, Ver 8. Cary, NC,USA.

34. ICPIN. 1993. A linear interpolation method for sublethal toxicity:The inhibition concentration (Icp) approach, Ver 2.0. U.S. En-vironmental Protection Agency, Duluth, MN.

35. Floate KD, Wardhaugh KG, Boxall ABA, Sherratt TN. 2005.Facial residues of veterinary parasites: Nontarget effects in thepasture environment. Ann Rev Entomol 50:153–179.

36. Strong L. 1993. Overview: The impact of avermectins on pas-tureland ecology. Vet Parasitol 48:3–17.

37. Jensen J, Krogh PH, Sverdrup LE. 2003. Effects of the antibac-terial agent tiamulin, olaquindox and metronidazole and the an-thelmintic ivermectin on the soil invertebrate species Folsomiafimetaria (Collembola) and Enchytraeus crypticus (Enchytraei-dae). Chemosphere 50:437–443.

38. Knacker T, van Gestel CAM, Jones SE, Soares AMVM, Schall-nass HJ, Forster B, Edwards CA. 2004. Ring-testing and field-validation of a terrestrial model ecosystem (TME): An instrumentfor testing potentially harmful substances. Conceptual approachand study design. Ecotoxicology 13:9–27.

39. Cortet J, Joffre R, Elmholt S, Coeurdassier M, Scheifler R, KroghPH. 2006. Interspecific relationships among soil invertebrates in-fluence pollutant effects of phenanthrene. Environ Toxicol Chem25:120–127.

40. Hamers T, Krogh PH. 1997. Predator–prey relationships in a two-species toxicity test system. Ecotoxicol Environ Saf 37:203–212.

41. Ignatowicz SP. 1974. Observation on the biology and develop-ment of Hypoaspis aculeifer Canestrini 1885 (Acarina, Gama-sides). Zool Pol 24:41–59.

42. Gunn A, Sadd JW. 1994. The effects of ivermectin on the survival,

behavior and cocoon production of the earthworm Eisenia fetida.Pedobiologia 38:327–333.

43. Halley BA, Jacob TA, Lu AYH. 1989. The environmental impactof the use of ivermectin: effects and fate. Chemosphere 18:1543–1563.

44. Svendsen TS, Hansen PE, Sommer C, Martinussen T, GrønvoldJ, Holter P. 2005. Life history characteristics of Lumbricus ter-restris and effects of the veterinary antiparasitic compounds iver-mectin and fenbendazole. Soil Biol Biochem 37:927–936.

45. Diao X, Jensen J, Hansen AD. 2007. Toxicity of the anthelminticabamectin to four species of soil invertebrates. Environ Pollut148:514–519.

46. Sommer C, Steffansen B, Overgaard Nielsen B, Grønvold J, VagnJensen KM, Brøchner Jespersen J, Springborg J, Nansen P. 1992.Ivermectin excreted in cattle dung after subcutaneous injectionor pour-on treatment: Concentration and impact on dung fauna.Bull Entomol Res 82:257–264.

47. Sommer C, Nielsen BO. 1992. Larvae of the dung beetle Onth-ophagus gazella F. (Col., Scarabaeidae) exposed to lethal andsublethal ivermectin concentrations. J Appl Entomol 114:502–509.

48. Fincher GT. 1992. Injectable ivermectin for cattle: Effects onsome dung-inhabiting insects. Environ Entomol 21:871–876.

49. Gover J, Strong L. 1995. The effects of ivermectin in ingestedcow dung on the mortality and oviposition of the dung fly Neo-myia cornicina (Diptera, Muscidae). Bull Entomol Res 85:53–57.

50. Gover J, Strong L. 1996. Effects of ingestion of dung containingivermectin on aspects of behaviour in the fly Neomyia cornicina.Physiol Entomol 21:51–58.

51. Oppel J, Broll G, Loffler D, Meller M, Rombke J, Ternes T. 2004.Leaching behaviour of pharmaceuticals in soil-testing systems:A part of an environmental risk assessment for groundwater. SciTotal Environ 328:265–273.

52. Suarez VH, Lifschitz AL, Sallovitz JM, Lanusse CE. 2008. Ef-fects of faecal residues of moxidectin and doramectin on theactivity of arthropods in cattle dung. Ecotoxicol Environ Saf (inpress).

53. European Food Safety Authority. 2005. Opinion of the ScientificPanel on Additives and Products or Substances Used in AnimalFeed on a request from the European Commission on the eval-uation of the coccidiostat COXIDIN� (monensin sodium). EFSAJ 283:1–53.

54. Herald F, Knapp FW, Brown S, Bradley NW. 1982. Efficacy ofmonensin as a cattle feed additive against the face fly and hornfly. J Anim Sci 54:1128–1131.

55. Carlson JC, Mabury SA. 2006. Dissipation kinetics and mobilityof chlortetracycline, tylosin, and monensin in an agricultural soilin Northumberland County, Ontario, Canada. Environ ToxicolChem 25:1–10.

56. European Food Safety Authority. 2005. Opinion of the ScientificPanel on Additives and Products or Substances Used in AnimalFeed on a request from the European Commission on the eval-uation of the coccidiostat COXIDIN� (monensin sodium). EFSAJ 283:Annex B.

single- and two-species tests to study effects of the anthelmintics ivermectin and morantel and the...

Documents