Integr Envir Assess Manag - 2010 - Liebig - Environmental Risk Assessment of Ivermectin A Case Study

Integrated Environmental Assessment and Management — Volume 6, Supplement 1—pp.
567–587
ß 2010 SETAC 567
Environmental Risk Assessment of Ivermectin: A Case Study

Markus Liebig,y * Álvaro Alonso Fernandez,z Elke Blübaum-Gronau,§ Alistair Boxall,k Marvin Brinke,#
Gregoria Carbonell,z Philipp Egeler,y Kathrin Fenner,yy,zz Carlos Fernandez,z Guido Fink,§ Jeanne Garric,§§
Bent Halling-Sørensen,kk Thomas Knacker,y Kristine A Krogh,kk Anette Küster,## Dirk Löffler,§ Miguel Ángel
Porcel Cots,z Louise Pope,k Carsten Prasse,§ Jörg Römbke,y Ines Rönnefahrt,## Manuel K. Schneider,yy
Natascha Schweitzer,y José V Tarazona,z Thomas A Ternes,§ Walter Traunspurger,# Anne Wehrhan,## and
Karen Duisy
y ECT Oekotoxikologie GmbH, Böttgerstrasse 2-14, D-65439 Flörsheim/Main, Germany
z Instituto Nacional de Investigación y Tecnologı́a Agraria y Alimentaria (INIA), Ctra La Coruña km 7.5, E-28040 Madrid, Spain
§ Federal Institute of Hydrology, Am Mainzer Tor 1, D-56068 Koblenz, Germany
k Central Science Laboratory (CSL), Sand Hutton, York YO41 1LZ, United Kingdom
#Department Animal Ecology, University of Bielefeld, Morgenbreede 45, D-33615 Bielefeld, Germany
yy Swiss Federal Institute of Aquatic Science and Technology (Eawag), Überlandstrasse 133, CH-8600 Dübendorf, Switzerland
zz Institute of Biogeochemistry and Pollutant Dynamics (IBP), ETH Zürich, Universitätsstrasse 16, CH-8092 Zürich, Switzerland
§§Cemagref, UR BELY, Laboratory of Ecotoxicology, 3 bis, Quai Chauveau, CP 220, F-69336 Lyon Cedex 09, France
Special Series
kk Faculty of Pharmaceutical Sciences, Copenhagen University, Universitetsparken 2, DK-2100 Copenhagen, Denmark
##Federal Environment Agency (UBA), Postfach 1406, D-06813 Dessau, Germany
(Submitted 11 May 2009; Returned for Revision 8 July 2009; Accepted 14 April 2010)
ABSTRACT
The veterinary parasiticide ivermectin was selected as a case study compound within the project ERAPharm (Environmental
Risk Assessment of Pharmaceuticals). Based on experimental data generated within ERAPharm and additional literature data, an
environmental risk assessment (ERA) was performed mainly according to international and European guidelines. For the
environmental compartments surface water, sediment, and dung, a risk was indicated at all levels of the tiered assessment
approach. Only for soil was no risk indicated after the lower tier assessment. However, the use of effects data from additional 2-
species and multispecies studies resulted in a risk indication for collembolans. Although previously performed ERAs for
ivermectin revealed no concern for the aquatic compartment, and transient effects on dung-insect populations were not
considered as relevant, the present ERA clearly demonstrates unacceptable risks for all investigated environmental
compartments and hence suggests the necessity of reassessing ivermectin-containing products. Based on this case study,
several gaps in the existing guidelines for ERA of pharmaceuticals were shown and improvements have been suggested. The
action limit at the start of the ERA, for example, is not protective for substances such as ivermectin when used on intensively
reared animals. Furthermore, initial predicted environmental concentrations (PECs) of ivermectin in soil were estimated to be
lower than refined PECs, indicating that the currently used tiered approach for exposure assessment is not appropriate
for substances with potential for accumulation in soil. In addition, guidance is lacking for the assessment of effects at higher
tiers of the ERA, e.g., for field studies or a tiered effects assessment in the dung compartment. Integr Environ Assess Manag
2010;6:567–587. ß 2010 SETAC
Keywords: Environmental risk assessment Fate and effects assessment Parasiticides Tiered approach Veterinary
pharmaceuticals
INTRODUCTION bacteria or invertebrates (Boxall et al. 2004). In this respect

The potential risk of veterinary medicinal products (VMPs) they are very similar to pesticides and biocidal products.
for the environment raised concern much earlier than that of There are even examples in which the same active substance
human medical products (HMPs). For example, the impact of is used for both purposes, i.e., as pesticide and VMP (e.g.,
parasiticides on the survival of dung beetles was studied more deltamethrin). Therefore, similar environmental problems are
than 30 y ago (Blume et al. 1976). VMPs often reach soils likely to occur for VMPs as for pesticides. Target (e.g., blow
more directly than HMPs, because VMPs such as endo- and flies or ascaricid roundworms) and nontarget organisms (e.g.,
ectoparasiticides are regularly applied to pasture animals and dung flies or saprophagous nematodes) can belong to the
intensively reared livestock. Residues can reach soils through same taxonomic groups, dipterans and nematodes, respec-
3 main exposure routes: directly via feces, indirectly via tively. Hence, the respective substances are likely to affect not
spread manure or through wash-off from topically applied only target but also nontarget organisms. The main difference
products (Halling-Sørensen et al. 1998). VMPs often act as between pesticides and VMPs is that the latter are often
biocides; i.e., they specifically act on target organisms such as excreted as a mixture of metabolites and parent compound,
whereas pesticides are released directly to the environment as
* To whom correspondence may be addressed: m-liebig@ect.de parent compound (Halling-Sørensen et al. 1998).
Published online 23 April 2010 in Wiley InterScience Avermectins are an important group of VMPs in terms of
(www.interscience.wiley.com).
both their widespread use and their potential environmental
risks (Campbell et al. 1983; Strong and Brown 1987). They
DOI: 10.1002/ieam.96
have been used in agriculture and horticulture for the
568 Integr Environ Assess Manag 6, 2010—M Liebig et al.
protection of fruits, cotton, vegetables, and ornamentals (Dybas addressed exposure and effects of ivermectin in the environ-
1989), because they are effective against a wide range of ment (e.g., Edwards et al. 2001; Boxall et al. 2004; Floate
nematodes, mites, and insects (Strong and Brown 1987; Õmura et al. 2005; Kolar and Kožuh Eržen 2006), but few were
2008). Avermectins are also used for treatment of river carried out according to standardized guidelines. Because of
blindness, i.e., onchocerciasis, in humans (Lindley 1987). its potential environmental effects and its economic impor-
However, the most extensive use of avermectins is in the tance, ivermectin was chosen as a case study compound
control of livestock parasites. The main route of excretion is via within the EU-funded project ERAPharm.
feces (Chui et al. 1990), which provides a microhabitat and In the European Union (EU), the evaluation of the
breeding ground for a very large number of invertebrate species, environmental risk of veterinary medicinal products within
on which avermectins are known to have deleterious effects. marketing authorization procedures has been discussed since
Avermectins are macrocyclic lactones isolated from the soil the mid-1990s (Koschorreck and Apel 2006), and a first
actinomycete Streptomyces avermitilis. The most well studied guidance document on how to perform the ERA was prepared
avermectin is ivermectin (consisting of 80% 22,23-dihy- by the European Medicines Agency in 1997 (EMEA 1997).
droavermectin B1a and 20% 22,23-dihydroavermectin B1b; From this document, the EU, the United States, and Japan
Figure 1), a synthetic derivative of the naturally occurring harmonized the ERA procedures and prepared 2 guidelines, of
avermectin B1. Ivermectin binds selectively and with high which the first focuses on exposure assessment (phase I; VICH
affinity to the ligand glutamate on the ligand-gated chloride 2000) and the second on a tiered risk assessment (phase II;
ion channels that occur in invertebrate nerve and muscle cells, VICH 2004). For the EU, additional guidance in support of the
causing irreversible opening of these channels (Rohrer and VICH guidelines is provided by EMEA (2008).
Arena 1995; Õmura 2002). Furthermore, ivermectin affects All fate and effect studies required for an ERA should be
g-aminobutyric acid (GABA)-related chloride ion channels performed according to international guidelines (e.g., OECD
occurring in the peripheral nervous system of invertebrates or ISO). In the ERAPharm project, all studies conducted with
and in the central nervous system of vertebrates (Duce and ivermectin fulfilled this criterion, except for the higher tier
Scott 1985). From a food safety perspective, the margin of studies, i.e., 2-species, multispecies, semifield, and field
safety for ivermectin is attributable to the facts that 1) studies, for which no guidelines are available. In addition,
mammals do not have glutamate-gated chloride channels, and reliable data from the scientific literature were used for the
2) the macrocyclic lactones have a low affinity for other ERA; data quality (reliability) was assessed according to
mammalian ligand-gated chloride channels and do not readily Klimisch et al. (1997). In general, only data considered as
cross the blood–brain barrier (Boelsterli 2003; Õmura 2008). reliable were used for the ERA. However, some ERAPharm
With over 5 billion doses sold worldwide since its market data included in this paper have recently been submitted for
introduction in the early 1980s, ivermectin has become the publication and are still being reviewed. Furthermore, it was
most widely used antiparasitic drug (Shoop and Soll 2002). It not in all cases feasible to perform the studies as required by
is used regularly as a parasiticide for cattle, pigs, sheep, horses, the underlying ERA procedure (VICH 2000, 2004; EMEA
and dogs (Campbell et al. 1983; Forbes 1993). Oral 2008). Hence, the presented ERA for ivermectin should be
applications tend to result in sharp excretion peaks, with partially regarded as preliminary.
most of the dose excreted over a few days. Peak elimination of Our present objectives are 1) to conduct an ERA for the
injectable or topical formulations usually occurs within 2 to 7 parasiticide ivermectin, mainly according to the current
d posttreatment, followed by a long tail that may sustain for guidelines for environmental impact assessment (VICH
more than 4 to 6 weeks, whereas peak elimination levels of 2000, 2004; EMEA 2008) but taking several species and
sustained-release formulations may occur over several weeks various routes of administration into account, and 2) to show
posttreatment (Floate et al. 2005). gaps and to propose improvements of the existing guidelines
Because of its very high acute toxicity to invertebrates (see, by integrating data derived from nonstandardized studies into
e.g., Blume et al. 1976; Campbell et al. 1983), an environ- higher tier risk assessment procedures.
mental risk assessment (ERA) for ivermectin was performed
as early as 1986 (USFDA 1986). Several studies have ENVIRONMENTAL RISK ASSESSMENT ACCORDING
TO VICH (2000, 2004) AND EMEA (2008)
In phase I, a number of questions concerning application
and properties of the VMP direct the ERA to the main
exposure scenarios, i.e., aquaculture, intensively reared, or
pasture animals (VICH 2000). Then, predicted environ-
mental concentrations (PEC) are estimated based on the dose
and frequency of the product applied. If the PEC exceeds the
trigger value of 100 mg/kg dry wt in soil for intensively reared
and pasture animals, studies on environmental fate and effects
on selected nontarget species have to be performed in phase II
(VICH 2004). For parasiticides used in treatment of pasture
animals, the PECsoil trigger is circumvented, and phase II
studies are necessarily independent of PECsoil. In phase II, the
environmental risk is characterized deterministically by
comparing the PECs with the predicted no effect concen-
trations (PNECs) in several environmental compartments.
According to the guidelines (VICH 2000, 2004), the initial
Figure 1. Chemical structure of ivermectin. ERA is based on worst-case assumptions (e.g., with regard to
Environmental Risk Assessment of Ivermectin—Integr Environ Assess Manag 6, 2010 569
applied dose, excretion, fate, and behavior in the environment), Table 1. Physicochemical properties of ivermectin
whereas for further refinements averaged values are used (e.g., (CAS 70288-86-7)
KOC), when data allow for averaging. In the present case study,
parameters such as DT50 and KOC were derived for different Molecular mass (g/mol) 874.7a
soils, which reflect several European regions and climatic
pKa Neutral at all pH values
conditions. Because these conditions vary considerably, it was
not assumed that data allow for averaging. Consequently, Melting point (8 C) 349.8b (est)
minimum and maximum PECs are shown, demonstrating the
Vapor pressure (Pa) <1.5 109c (m)
possible range of environmental exposure resulting from the
use of veterinary medicines containing ivermectin. Henry constant () 4.8 1026b (est)
Water solubility (mg/L) 4.0d (m), 4.1e (m), 2.0f (m)

PHASE I
According to our knowledge, ivermectin is not currently Log KOW () 3.2d (m)
used in marine aquaculture in Europe. Hence, this scenario is
not considered in the present case study. Predicted environ- Log KOC (L/kg) 3.6–4.4g (m)
mental concentrations of ivermectin in soil (PECsoil) were UV-visible absorption Maxima: 237, 245 and 253 nm
calculated for the intensively reared (IR) and pasture animal spectrum (subject to direct photolysis)c
scenarios (P), considering worst-case assumptions (EMEA
a
2008). All estimated initial PECsoil values were below the Referring to ivermectin consisting of 94% B1a and 2.8% B1b, which was used
action limit of 100 mg/kg dry wt (see Predicted environmental in most of the tests performed within ERAPharm.
b
U.S. EPI-Suite v.4.00 (2008).
concentrations section). However, because ivermectin is c
Halley, Nessel, et al. (1989).
administered as an endo- and ectoparasiticide to animals d
USFDA (1990). Dossier data, no details on experimental methods are avail-
reared on pasture, e.g., cattle and sheep, a phase II assessment able.
e
is required independent of the PECsoil (VICH 2004). Escher et al. (2008). Determined using a modified shake flask method
according to Avdeef et al. (2007).
Although not required by VICH (2000), a phase II assessment f
Escher et al. (2008). Intrinsic solubility determined using a mDISS ProfilerTM.
was also performed in this study for ivermectin administered g
Krogh et al. (2008).
to intensively reared animals. m ¼ Measured; est ¼ estimated.
PHASE II TIER A The limited mobility of ivermectin in soils justifies the

In phase II, the PECs for various environmental compart-
assumption of little potential for groundwater contamination.
ments are compared to the corresponding PNECs (VICH
Transformation of ivermectin in soil was investigated in
2004). If in phase II tier A, a compartment-specific PEC
ERAPharm according to OECD 307 (2002a) using non-
exceeds the organism-specific PNEC, an environmental risk is
labeled ivermectin. The results indicate that dissipation half-
indicated, and tier B testing for the specific compartment
lives (DT50) in soil can be rather variable depending on soil
including the organisms of concern is required. Phase II tier A
type, sorption capacity, temperature, and oxygen availability
assessment relies on a base set of data on physicochemical
(Krogh et al. 2009). The highest DT50 of 67 d was derived
properties (Table 1), on environmental fate, and on effects
with a simple first-order model for natural soil at 208C under
determined in single-species tests under laboratory condi-
aerobic conditions (Table 3). This DT50 was used as a worst-
tions.
case value in the exposure assessment. Within the study of
Krogh et al. (2009), 2 transformation products of ivermectin
were identified in soil, a monosaccharide and an aglycone of
Environmental fate ivermectin (22,23-dihydroavermectin B1 monosaccharide and
In ERAPharm, sorption was determined mostly according 22,23-dihydroavermectin B1 aglycone; our observations).
to OECD 106 (OECD 2000) for artificial and 2 natural loamy However, the transformation products were quantified at
soils using 3H-labeled and nonlabeled ivermectin (Table 2) levels <10% of the parent compound, so no transformation
(Krogh et al. 2008). Halley, Jacob, et al. (1989) studied products were considered in the present ERA.
sorption of ivermectin in a clay loam and a silty loam soil Literature data from mostly nonstandardized biodegradation
(Table 2). Equilibrium distribution was reached within 48 h tests indicate a broad range of DT50 values resulting in classi-
(Krogh et al. 2008) and 16 h (Halley, Jacob, et al. 1989). The fications ranging from slightly to moderately persistent in soil
estimated Kd values (average values of sorption and 2 (DT50 ¼ 14–56 d) to slightly to very persistent in mixtures of
desorption steps) from the latter experiments were 227 and soil and manure or feces (DT50 ¼ 7–217 d; Boxall et al. 2002).
333 L/kg, corresponding to KOC values of 1.48 104 and Halley, Jacob, et al. (1989) investigated the aerobic trans-
1.57 104 L/kg, indicating strong sorption (Halley, Jacob, formation of ivermectin in soil–feces mixtures and determined
et al. 1989). DT50 values of 93 d and 240 d, depending on soil type and
In soil column experiments with 2 soils containing 2.3 and mode of application. Reports of low ivermectin persistence in
6.3% organic carbon content, no ivermectin was detected in manure following summer or dry conditions might be an
the leachate (Oppel et al. 2004), whereas in another study, artefact resulting from reduced ivermectin extraction efficiency
27% to 48% of the applied 3H radioactivity was leached as at low moisture content of the solid matrix (Pope 2010).
transformation products, and 39% to 49% remained in the top Degradation of ivermectin in water–sediment systems was
5 cm of the soil column (Halley, Jacob, et al. 1989). The investigated within ERAPharm according to OECD 308
identity of this strongly sorbed fraction remained undeter- (2002b) using natural sediment containing 4.5% total organic
mined but was assumed to be mostly the parent substance. carbon (TOC), with resulting compartment-specific degrada-
Table 2. Soil parameters, sorption/desorption properties, and organic carbon normalized adsorption coefficients (KOC) for 5 different soils
Soil type pH fom foc Kd (L/kg) Kdes (L/kg) KOC (L/kg) Log KOC
Artificial (OECD)a 6.0 0.047 0.0273 109 141–246 4.00 103 3.6
a 4
York, UK 6.3 0.0265 0.0154 396 54–201 2.58 10 4.4
Madrid, Ea 8.7 0.0077 0.0045 57 28–56 1.28 104 4.1

b c 4
Newton, USA 5.5 0.039 0.0226 333 n. d. 1.47 10 4.2
Fulton, USAb 6.3 0.025 0.0145 227c n. d. 1.57 104 4.2
Italicized values were used for best- and worst-case exposure assessment. fom ¼ Fraction of organic matter; foc ¼ fraction of organic carbon (converted from Fom
according to Halley, Jacob, et al. 1989); Kd ¼ measured soil–water distribution coefficient; Kdes ¼ measured desorption coefficient; KOC ¼ organic carbon
normalized adsorption coefficient calculated according to KOC ¼ Kd/foc; n.d. ¼ not determined.
a
Soils investigated within ERAPharm (Krogh et al. 2008), mostly according to OECD 106 (only 0.5 g soil was used and Freundlich isotherms were determined only
for 1 soil type).
b
Clay loam and silty clay loam (Halley, Jacob, et al. 1989); KOC values were recalculated according to Halley, Jacob, et al. (1989): Kd 100/(fom/1.72).
c
Soil/0.01 M CaCl2 partition coefficient, average of sorption and 2 desorption steps (Halley, Jacob, et al. 1989).
tion half-lives (t1/2) as shown in Table 3 and an estimated et al. 2005; Prasse et al. 2009). In a long-term outdoor aquatic
dissipation half-life (DT50) in water of <0.25 d (Prasse et al. mesocosm study (265 d) with ivermectin using natural water
2009). Löffler et al. (2005) also investigated the fate of and sediments, a DT50 of 4 d was derived for the water
ivermectin in water–sediment systems. The authors found a phase. However, no DT50 for sediment could be determined,
dissipation half-life (DT50) of 15 d for the whole system because after reaching a steady state, no dissipation of
containing natural sediment with 1.4% TOC; the DT50 from ivermectin from the sediment was discernible until the end
the water phase was estimated to be 2.9 d (Löffler et al. of the study (Sanderson et al. 2007).
2005). The sediment–water distribution coefficients (Kd Ivermectin is hydrolytically unstable both in acidic and in
sediment) of ivermectin were 160 and 11.7 L/kg, corresponding basic solution, being most stable at a pH of 6.3 (Fink 1988).
to KOC values of 3550 and 1172 L/kg, respectively (Löffler Data on hydrolysis in environmental matrices were not
Table 3. Transformation of ivermectin in soils and aquatic sediments
Type of study Value Reference
Transformation in soil (OECD 307)a
Dissipation (Madrid soil) DT50 16 d Krogh et al. (2009)
DT90 54 d Krogh et al. (2009)
Dissipation (York soil) DT50 67 d Krogh et al. (2009)

b
DT90 222 d Krogh et al. (2009)
Dissipation (artificial soil) DT50 458 db Krogh et al. (2009)

b
DT90 1520 d Krogh et al. (2009)
Transformation in water–sediment systems (OECD 308)
Dissipation: DT50 (water) <0.25 d Prasse et al. (2009)
Dissipation: DT50 (whole system) 127 d Prasse et al. (2009)
Degradation: t1/2 (water) 30 d Our calculations based

on Prasse et al. (2009)
Degradation: t1/2 (sediment) 130 d Our calculations based

Degradation: t1/2 (whole system) 87 d Our calculations based

Dissipation: DT50 (water) 2.9 d Löffler et al. (2005)
Dissipation: DT50 (whole system) 15 d Löffler et al. (2005)
Italicized values were used for best- and worst-case exposure assessment.
a
Calculated with simple first order model (OECD 2002a); conditions: aerobic at 208C (Krogh et al. 2009).
b
Values above 120 d are extrapolated; the last sampling took place at day 120.
available in the scientific literature. The photolytic half-life of pastures or arable land after application of manure from
ivermectin determined in a thin, dry film exposed to direct treated animals. The sediment compartment may be con-
sunlight was approximately 3 h (Halley, Jacob, et al. 1989). taminated via transfer from surface waters into sediments or
Photoinduced reactions are thus anticipated to influence the sedimentation of eroded material from pastures or arable
fate of ivermectin in the aquatic environment. Studies on land.
photolysis and hydrolysis might be required by regulatory Because of its high affinity for soil and particulate matter,
authorities based on expert judgement. However, the results neither leaching nor runoff was assumed to be a major source
of a long-term outdoor aquatic mesocosm study (Sanderson for contamination of freshwater ecosystems with ivermectin
et al. 2007) with ivermectin using natural water and sediment (Kövecses and Marcogliese 2005). However, the transport of
suggest that both processes play a minor role, insofar as sorbed ivermectin with eroded soil might be important. The
ivermectin dissipates rapidly from the water phase into the risk of soil translocation from erosion is highest when crop
sediment. coverage is lowest, i.e., in fall after harvesting or in spring
before seeding. The postharvest (and preseeding) period with
a high erosion risk coincides with the time when large
Predicted environmental concentrations numbers of animals are treated with ivermectin and farmers
Ivermectin may enter the terrestrial compartment via are allowed to spread manure (Kövecses and Marcogliese
spreading of manure from intensively reared animals on 2005). It may in some regions also coincide with the time of
arable land or by excretion of dung by animals on pastures. intensive rainfall events, initiating soil erosion.
Likewise, it can be released directly to surface water via In Table 4, the initial PECs are shown for those environ-
treated animals (e.g., cattle) standing in shallow water bodies. mental compartments involved in environmental fate and
Indirect entry into water might occur via leaching from behavior processes relevant for ivermectin. The initial PECs
contaminated soil into groundwater or via runoff from were calculated according to the total residue approach, in
Table 4. Initial PECs of ivermectin in different environmental compartments: soil (PECsoil), groundwater (PECgw), surface water (PECsw), and
dung (PECdung), calculated according to EMEA (2008)
Compartment Unit PEC Remark
PECsoil initial (IR)a,b mg/kg dry wt 2.61/6.08c Weaner pig (<25 kg), H ¼ 1
c
0.63/1.47 Sow with litters, H ¼ 1
a c
PECsoil initial (P) mg/kg dry wt 0.84/2.09 Beef cattle
0.33 Pony
d c
PECsoil plateau (IR) mg/kg dry wt 2.67/6.22 Weaner pig (<25 kg), H ¼ 1
c
0.64/1.50 Sow with litters, H ¼ 1
PECsoil plateau (P)d mg/kg dry wt 0.86/2.14c Beef cattle
0.34 Pony
PECgw initial (IR) ng/L 3.3–21.5e Weaner pig (PECgw ¼ PECporewater)
PECgw initial (P) ng/L 0.5–7.4e Beef cattle (PECgw ¼ PECporewater)

e
PECsw initial (IR) ng/L 0.1–7.2 Sow with litters weaner pig
(PECsw ¼ 1/3 PECporewater)
PECsw initial (P) ng/L 0.2–2.5e Beef cattle (PECsw ¼ 1/3 PECporewater)
f c
PECsw initial (P; d.e. ) ng/L 209/523 Beef cattle
83 Pony
c,g
PECdung initial (P) mg/kg dung 5.08/12.69 Beef cattle
fresh wt
4.8g Horse
PECs are shown only for those species with highest and lowest values for the respective compartment and scenario. IR ¼ intensively reared animals; P ¼ pasture
animals; H ¼ housing factor (fraction of the year in which the animals are kept in house).
a
Initial PECssoil at 5 cm mixing depth.
b
Assuming the EU nitrogen spreading limit of 170 kg N/(ha y).
c
Calculated with minimum/maximum dose.
d
PECsoil at steady state considering degradation properties and accumulation of ivermectin in soil.
e
Range from maximum best-case to maximum worst-case PEC calculated with maximum and minimum KOC value, respectively (Table 2).
f
PECsw calculated for the specific scenario of direct excretion (d.e.) into surface waters from pasture animals.
g
Values based on dry wt: 28.6/71.5 for beef cattle and 27.0 for horse (conversion factor fresh wt/dry wt ¼ 5.63; our results).
which it is assumed that 100% of the total dose administered directly into the stream (EMEA 2008). For other specific P
during the treatment is released to the environment (EMEA scenarios proposed by EMEA (2008), e.g., runoff from
2008). Calculation of PECs is based on different types of contaminated hard standing areas, neither a model to
dosages (0.1–0.5 mg/kg body wt) and application frequencies calculate the specific PEC nor relevant data were available.
(1, 2, or 7 applications) to several productive livestock For the dung compartment, initial PECs for application to
species. This information was compiled from summaries of all target animal species were between 4.8 and 8.0 mg/kg
product characteristics for ivermectin-containing products dung fresh wt, except for beef cattle (12.69 mg/kg dung fresh
(Chanectin1, Diapec1, Ecomectin1). In addition to the wt). Halley, Nessel, et al. (1989) derived PECs for ivermectin
maximum PEC values as requested according to EMEA in dung and soil following administration of ivermectin to
(2008), minimum PEC values are indicated. various livestock in feedlots or on pasture. In contrast to the
For the soil compartment, a range of PECs was derived for total-residue approach proposed by EMEA (2008), they
the IR and P scenario, with a minimum of 0.33 mg/kg dry wt assumed constant excretion of the applied dose over a feedlot
and a maximum of 6.08 mg/kg dry wt, estimated for ponies for a period up to 120 to 168 d. Ivermectin concentrations in
and weaner pigs, respectively (Table 4). For persistent feces were estimated to be 18 to 19 mg/kg dung fresh wt for
compounds (DT90soil >1 y), accumulation in soil after swine, sheep, and cattle. Assuming manure application under
application of manure during successive years is possible, good agricultural practice and 15-cm plowing depth resulted
and, hence, a PECsoil plateau at steady state should be in a PECsoil of 0.2 mg/kg dry wt for intensively reared cattle
calculated according to EMEA (2008). Although not required and swine. The estimated application rates for sheep and
for ivermectin (DT90 ¼ 222 d; see Table 3), the worst-case cattle dung on pasture were 0.013 and 0.016 mg ivermectin/
PECsoil plateau was calculated. Because this value, 6.22 mg/kg m2, respectively (Halley, Nessel, et al. 1989). However,
dry wt, is only slightly above the initial PECsoil of 6.08 mg/kg Fernandez et al. (2009) and Lumaret et al. (2007) measured
dry wt, it was not used further in the ERA. ivermectin concentrations of 145 mg/kg dung fresh wt and
The concentrations of ivermectin in groundwater and approximately 250 mg/kg dung fresh wt in cattle dung at the
surface water were estimated based on the PECsoil, assuming excretion peak, which are much higher than the value
that the concentration in groundwater equals the concen- estimated by Halley, Nessel, et al. (1989).
tration in soil porewater at a mixing depth of 20 cm.
PECporewater was calculated assuming sorption equilibrium
of ivermectin between soil and porewater, characterized by Aquatic short-term effect studies
Kd or KOC (Table 2). Using the lowest and highest log KOC The base set data according to EMEA (2008) on short-term
values (3.6 and 4.4) and the minimum and maximum PECsoil effects of ivermectin to fish, Daphnia, and algae from the
(0.33 and 6.08 mg/kg dry wt), the predicted groundwater literature was supplemented with data derived from ERA-
concentrations range from 0.5 to 21.5 ng/L (Table 4). The Pharm (Garric et al. 2007; Table 5). Within ERAPharm, a
initial PECsw is assumed to be one-third of the soil porewater growth inhibition test with the green alga P. subcapitata
concentrations (EMEA 2008) resulting in initial PECsw values exposed to ivermectin was performed according to OECD
from 0.1 to 7.2 ng/L. 201 (2002c). EC50 for yield and growth rate was >4.0 mg/L,
The specific P scenario of direct excretion by pasture cattle and NOEC was 391 mg/L (Garric et al. 2007). Ten Daphnia
via urine or feces into surface water takes into account a immobilization tests were performed according to OECD 202
standard pasture of 1 ha containing a shallow, slow-flowing (2004a). To avoid photodegradation, these tests were con-
ditch covering 1% of the area. It is assumed that pasture ducted in the dark. EC50 values ranged from 1.2 to 10.7 ng/L
animals excrete 1% of the total dose administered within 1 d (mean value 5.7 ng/L; Garric et al. 2007). These values are
Table 5. Phase II tier A aquatic short-term effect studies
Test organism Test method Effect concentration Reference

a,b
Pseudokirchneriella OECD 201 (2002c) EC5072 h, yield, growth rate >4 mg/L Garric et al. (2007)
subcapitata (green alga)
LOEC72 h, yield, growth rate ¼ 1.25 mg/La

a
NOEC72 h, yield, growth rate ¼ 391 mg/L
Daphnia magna (crustacean) OECD 202 (2004a) EC5048 h, immobility ¼ 1.2–10.7 ng/Lc Garric et al. (2007)
c
Mean EC5048 h ¼ 5.7 ng/L (n ¼ 10)
USEPA 660/3-75-009 (1975) LC5048 h ¼ 25 ng/La Halley, Jacob, et al. (1989)
Oncorhynchus mykiss (fish) USEPA 660/3-75-009 (1975) LC5096 h ¼ 3.0 mg/La Halley, Jacob, et al. (1989)
a
Salmo salar (fish) Acute toxicity test (juvenile fish) LC5096 h ¼ 17 mg/L Kilmartin et al. (1996)
Results of the most sensitive tests (italicized) were used for the risk characterization.
a
Based on nominal concentrations.
b
According to VICH (2004), the EC50 is used for risk characterization in phase II tier A.
c
Based on measured concentrations.
slightly below the LC50 of 25 ng/L derived for D. magna by Some EU authorities require information on the toxicity to
Halley and colleagues (Halley, Jacob, et al. 1989; Halley, nontarget arthropods for parasiticides for the IR scenario, so
Nessel, et al. 1989). As far as is known from scientific literature, collembolan reproduction tests were performed according to
acute effects of ivermectin on fish occur in the lower micro- ISO 11267 (ISO, 1999). As expected when considering the
grams-per-liter range, with Oncorhynchus mykiss as the most mode of action of ivermectin and the taxonomic relationship
sensitive species. In addition to the standard base set of acute- of collembolans to the target organisms, the tests revealed a
effects data for algae, Daphnia, and fish, acute-effects data are high sensitivity as shown by the NOEC of 0.3 mg/kg dry wt
available for estuarine and marine crustaceans, mollusks, and (Jensen et al. 2003; Römbke, Krogh, et al. 2010). Earthworms
other invertebrates. Overall, crustaceans are the most sensitive and other oligochaetes were less sensitive, with NOECs in the
taxonomic group, showing effect concentrations in the lower milligrams-per-kilogram range.
nanograms-per-liter range (see, e.g., Davies et al. 1997; Grant Because ivermectin is used to treat livestock on pasture,
and Briggs 1998; Garric et al. 2007). tests with dung beetles and dung flies are required in tier A.
Table 7 summarizes the results of dung fly and dung beetle
tests performed within ERAPharm as well as studies
Terrestrial effect studies described in the literature. The high sensitivity of Musca
Results of the terrestrial tests from ERAPharm and the autumnalis to ivermectin was confirmed in a ring test
literature are summarized in Table 6. As required by VICH performed to validate the OECD draft guideline (Römbke,
(2004), an earthworm reproduction test according to OECD Alonso, et al., 2010), where a mean EC50 of 4.65 mg/kg dung
220/222 (2004b, 2004c) was performed, resulting in an EC50 fresh wt was determined. In the literature, effect concen-
of 5.3 mg/kg dry wt and an NOEC of 2.5 mg/kg dry wt trations of 0.5 mg/kg dung fresh wt were reported for the
(Römbke, Krogh, et al. 2010). Because endo- and ectopar- yellow dung fly Scathophaga stercoraria when studying
asiticides are not considered to be toxic for plants and morphological changes in adults (Strong and James 1993).
microorganisms and the trigger value of 100 mg/kg for PECsoil However, these specific endpoints are difficult to assess and
given in phase I was not exceeded by ivermectin, neither a were not used for risk characterization. With LC50 values of
nitrogen transformation nor a plant test is required according 100 and 176 mg/kg dung fresh wt, the dung beetle Aphodius
to VICH (2004). constans reacted less sensitively to ivermectin than dung flies
Table 6. Phase II tier A terrestrial effect studies with soil organisms
Test organism Test method Effect concentrationa Reference
Eisenia fetida (earthworm) OECD 222 (2004c) NOEC28 d, biomass ¼ 5.0 mg/kg dry wt Römbke, Krogh,
(artificial soil, TOC 3.6%) et al. (2010)
NOEC56 d, reprod. ¼ 2.5 mg/kg dry wt
EC5056 d, reprod. ¼ 5.3 mg/kg dry wt
Eisenia fetida (earthworm) Subchronic earthworm NOEC28 d, biomass ¼ 12 mg/kg dry wt Halley, Jacob,
toxicity test (artificial soil) et al. (1989)
LC5028 d ¼ 315 mg/kg dry wt
Eisenia fetida (earthworm) OECD 207 (1984) (artificial soil) NOEC14 d, biomass ¼ 4 mg/kg dry wt Gunn and Sadd
(1994)
LC5014 d ¼ 15.8 mg/kg dry wt

b
Enchytraeus crypticus (potworm) ISO 16387 , (field soil: TOC 1.6%) NOEC28 d, reprod. ¼ 3.0 mg/kg dry wt Jensen et al. (2003)
EC5028 d, reprod. ¼ 36 mg/kg dry wt
LC5028 d >300 mg/kg dry wt
Folsomia candida (collembolan) ISO 11267 (1999) NOEC28 d, reprod. ¼ 0.3 mg/kg dry wt Römbke, Krogh,
(artific. soil: TOC 3.6%) et al. (2010)
Folsomia fimetaria (collembolan) ISO 11267 (1999) (field NOEC28 d, reprod. ¼ 0.3 mg/kg dry wt Jensen et al. (2003)
soil: total carbon 1.6%)
LC5028 d ¼ 8.4 mg/kg dry wt
a
Effect concentrations refer to nominal concentrations.
b
The test was performed according to a slightly modified method described by Römbke and Moser (1999) published as ISO 16387 (2004).
Table 7. Phase II tier A terrestrial effect studies with dung organisms
Musca autumnalis OECD (2008a) EC5021 d, emergence rate ¼ 4.65 mg/kg dung fresh wt Römbke, Barrett,
(dung fly) et al. (2010)
Scathophaga stercoraria OECD (2008a) LC5028 d ¼ 20.9 mg/kg dung fresh wt Römbke et al. (2009)
(dung fly)
NOEC28 d, development time ¼ 0.84 mg/kg dung fresh wt
Specific test design LC5048 h, larvae ¼ 36 mg/kg dung fresh wt Strong and James (1993)
(acute toxicity)
EC503–4 w., emergence ¼ 1.0 mg/kg dung fresh wt
Aphodius constans OECD draft (2009) LC5021 d ¼ 176 mg/kg dung fresh wt Hempel et al. (2006)
(dung beetle)
LC5021 d ¼ 880 mg/kg dung dry wt
NOEC21 d, larval survival ¼ 320 mg/kg dung dry wt
Aphodius constans OECD draft (2009), LC5021 d ¼ 100 mg/kg dung fresh wtb Lumaret et al. (2007)
(dung beetle) modified
LC5021 d ¼ 590 mg/kg dung dry wt
a
All effect concentrations refer to nominal concentrations.
b
Instead of spiked dung as recommended in OECD (2009), dung from treated cattle was used. The resulting EC50 was, thus, not used for phase II tier A risk
characterization.
(Hempel et al. 2006; Lumaret et al. 2007). The LC50 of LC50. This value is within the range of the initial PECsw. The
176 mg/kg dung fresh wt was used for the ERA. A lower LC50 RQ using the worst-case PECsw for the IR scenario is above
of 100 mg/kg dung fresh wt was derived with dung from the threshold of 1, indicating a risk for freshwater fish. For the
treated cattle (Lumaret et al. 2007). This approach is not specific P scenario assuming direct excretion from the treated
recommended by OECD (2009) but is considered to be animals into surface waters, the initial PECsw values are
appropriate for higher tier testing, in that it reflects a more higher, thus also indicating a risk. The most sensitive aquatic
realistic exposure scenario. species is the crustacean D. magna, with a mean EC50 of
Large numbers of additional tests with dung fauna species 5.7 ng/L (Table 5), which was used to derive the PNEC of
were performed, but most of them have limited value for a 5.7 pg/L. For all scenarios, the RQs indicate a high risk for
quantitative ERA, because NOEC or ECx values were not aquatic invertebrates (Table 8).
determined. In particular, only information on mortality in To derive PNECs for the terrestrial compartment, the
relation to the age of the dung is given in tests with treated EC50 of the plant and the LC50 of the dung organism
dung as test substrate. Because the concentration of ivermectin toxicity tests are divided by an AF of 100, whereas the
can hardly be related to the observed effects, these data (e.g., NOECs from the chronic earthworm and collembolan
NRA 1998; Steel and Wardhaugh 2002) are not taken into toxicity tests are divided by an AF of 10 (Table 9). The most
consideration for the risk assessment. In a test with dung-living sensitive endpoint for soil organisms was collembolan
nematode species, an NOEC of 3.0 mg/kg dung fresh wt was reproduction. However, the risk quotient between 0.01 and
determined (Grønvold et al. 2004), which is higher than the 0.48 did not indicate a risk for soil arthropods. For dung
values found for dung flies and beetles, although both insects beetles, the LC50 of 176 mg/kg dung fresh wt derived from a
and nematodes belong to the target organisms of ivermectin. test with spiked dung was used for the risk assessment. The
resulting RQs range from 2727 to 317 250, indicating a high
risk for dung organisms (Table 9).
Risk characterization According to VICH (2004), a risk characterization for
Based on the data shown in Tables 4 and 5, the risk sediment is required when the initial RQ for aquatic inverte-
quotient (RQ), i.e., the ratio of initial PEC to PNEC, for the brates is 1, which is the case for ivermectin (Table 8). In
aquatic compartment was determined. According to VICH applying the equilibrium partitioning model (EMEA 2008),
(2004), an assessment factor (AF) of 1000 was applied to the PNECsediment was 0.0012 and 0.0074 mg/kg dry wt when
acute effect concentrations for daphnids (EC50) and fish using the lowest and the highest KOC, respectively (Table 2).
(LC50) and an AF of 100 to the EC50 for algae in order to Likewise, the initial estimation of PECsediment was based on
derive the PNECs (Table 8). the lowest KOC and minimum PECsw as well as on the highest
Ivermectin is unlikely to present a risk for freshwater algae. KOC (Table 2) and maximum PECsw (Table 4) of ivermectin.
For fish, a PNEC of 3 ng/L was derived based on the lowest The resulting RQs shown in Table 10 are far above 1.
Table 8. Phase II tier A risk assessment for ivermectin in the aquatic compartment
PECsw
Species Effect concentration AF PNEC (best/worst case) RQ (best/worst case)
Pseudokirchneriella EC50 >4 mg/L 100 >40 mg/L 0.1/7.2 (IR) 2.5 106/1.8 104 (IR)
subcapitata
0.2/2.5 (P) 5.0 106/6.3 105 (P)
83/523 (P; d.e.) 2.1 103/1.3 102 (P; d.e.)
ng/L
Daphnia magna EC50 ¼ 5.7 ng/L 1000 0.0057 ng/L 18/1263 (IR)
35/439 (P)
14561/91754 (P; d.e.)
Oncorhynchus mykiss LC50 ¼ 3.0 mg/L 1000 3.0 ng/L 0.03/2.4 (IR)
0.07/0.8 (P)
27.7/174 (P; d.e.)
Values in boldface indicate a risk. AF ¼ assessment factor; PNEC ¼ predicted no effect concentration; PECsw ¼ initial predicted environmental concentration in
surface waters (maximum best-case and maximum worst-case values) for intensively reared (IR) and pasture (P) animal scenarios (see Table 4); RQ ¼ risk quotient
(PEC to PNEC ratio); d.e. ¼ direct excretion scenario.
Consequently, refinement of PECsediment and effects testing of a single dose of 200 mg/kg body wt. The peak of excretion
using sediment-dwelling organisms and spiked sediment is was observed 5.6 days postinjection, with 872 mg/kg dung dry
required (VICH 2004; EMEA 2008). wt (145 mg/kg dung fresh wt; Fernandez et al. 2009). During a
period of 31 d postinjection, 35% (10%) of the applied dose
was excreted as parent compound. Based on the daily dung
Refinement of PEC estimation production of 3.8 kg dry wt (our measurements made within
Exposure assessment can be refined by taking into account the project ERAPharm), the fraction of the total applied dose
metabolism, excretion pattern, and biodegradation of the at the peak of excretion was 3.31%. These experimental data
VMP in aquatic systems, soil, and dung. Fernandez et al. on metabolization of ivermectin in cattle dung correspond
(2009) studied metabolism of ivermectin in cattle dung well to investigations by Cook et al. (1996), who measured
excreted over a period of 31 d after subcutaneous application excretion peaks between 2.38 and 1.1 mg/kg dung dry wt on
Table 9. Phase II tier A risk assessment for ivermectin in the terrestrial compartment
Effect PEC RQ
Species concentrations AF PNEC (best/worst case) (best/worst case)
Soil Vicia sativa, EC50 >10 mg/kg 100 100mg/kg 0.63/6.08 (IR) 0.006/0.06 (IR)
Triticum aestivum soil dry wt soil dry wt
0.33/2.09 (P) mg/kg 0.003/0.02 (P)

soil dry wt
Eisenia fetida NOECreprod. ¼ 2.5 mg/kg 10 250 mg/kg 0.003/0.02 (IR)

soil dry wt soil dry wt
0.001/0.008 (P)
Folsomia candida NOECreprod. ¼ 0.3 mg/kg 10 30 mg/kg 0.02/0.20 (IR)

soil dry wt soil dry wt
0.01/0.07 (P)
Dung Musca autumnalis EC50emerg.rate ¼ 4.65 mg/kg 100 0.0465 mg/kg dung 4.8/12.7 (P) mg/kg 103 226/273 118 (P)
dung fresh wt fresh wt dung fresh wt
Aphodius constans LC50 ¼ 176 mg/kg 100 1.76 mg/kg 2727/7210 (P)
dung fresh wt dung fresh wt
Values in boldface indicate a risk. AF, PNEC, RQ as described for Table 8; PEC ¼ initial predicted environmental concentration in soil or dung for intensively reared
(IR) and pasture (P) animal scenarios (Table 4).
Table 10. Phase II tier A risk assessment for ivermectin in the sediment based on equilibrium partitioning (EMEA 2008)
PNECD. magna PNECsed (best/worst case) PECsed (best/worst case) RQ (best/worst case)
0.0057 ng/L 0.0074/0.0012mg/kg dry wt 0.02/9.25 (IR) 2.7/7708 (IR)
0.03/3.18 (P) 4.1/2650 (P)
16.7/675 (P; d.e.) 2257/562 500 (P; d.e.)
mg/kg dry wt
Values in boldface indicate a risk. PNECD. magna ¼ PNEC derived from acute toxicity to D. magna (Table 8); PNECsed ¼ predicted no effect concentration for
sediment organisms; PECsed ¼ initial predicted environmental concentration in sediment for intensively reared (IR) and pasture (P) animal scenarios derived by
equilibrium partitioning (EMEA 2008); RQ ¼ risk quotient (PEC to PNEC ratio); d.e. ¼ direct excretion scenario.
days 6 and 8 postinjection. Using a reverse isotope dilution the metamodel yielded values of <0.1 mg ivermectin/L for the
assay, Halley and colleagues (Halley, Jacob, et al. 1989; worst- and best-case scenario, running the PEARL model to
Halley, Nessel, et al. 1989) found that 39 to 44% of the total estimate groundwater concentrations was not considered
radioactivity in feces of 3H-ivermectin-treated steers was the necessary.
unaltered active ingredient. To estimate the long-term exposure concentrations in
Within ERAPharm, 2 potential metabolites of ivermectin surface water and sediment, FOCUS requires the degradation
were identified in cattle dung: 24-hydroxymethyl-H2B1a and half-lives t1/2 for ivermectin determined in the water–
300 -O-desmethyl-H2B1a (Pope 2010). These metabolites were sediment transformation study (OECD 308; Table 3).
also reported to be the most prominent in cattle and swine According to FOCUS (2006), the best-fit degradation rate
liver (Chiu et al. 1986, 1990; Halley et al. 1992). However, constants are kw ¼ 0.0229/d (corresponding to t1/2 water ¼ 30
the potential metabolites could not be quantified because of d) and ksed ¼ 0.0054/d (corresponding to t1/2 sediment ¼ 130 d).
time constraints on the preparation of the appropriate These data were used in different combinations together with
metabolite standards. According to the chromatograms, the the best-case (16 d) and worst-case (67 d) DT50 in soil (Table
amount of the metabolites was estimated to be less than the 3) to run the FOCUS models. The FOCUS shell SWASH was
amount of parent compound (Pope 2010). In addition, the used to run the 3 models (MACRO, PRZM, and TOXSWA)
more polar degradation products of ivermectin (monosac- necessary to calculate contamination of surface water and
charide and aglycone), as detected as transformation products sediment resulting from runoff and drainage. From the
in soil (see above), were shown to be less toxic to daphnids combination of the different FOCUS scenarios (e.g., drainage,
than the parent compound (Halley, Jacob, et al. 1989). runoff) with different water bodies (e.g., pond, stream), 14
Therefore, the PEC refinement taking metabolization and scenarios were identified, for which concentration courses
excretion data into account was performed based on the were calculated for a period of 1 y. Maximum annual
percentage of excreted parent compound (35%). concentrations were 0.77 and 6.2 ng/L in surface water and
For pasture animals directly excreting into surface waters 0.17 and 0.25 mg/kg wet wt in sediment, assuming best- and
(P; d.e.), refined PECsw and PECsediment were calculated by worst-case sorption and degradation, respectively (Table 11).
taking into account sorption and distribution properties of For better comparison with PNECs derived from chronic-
ivermectin and assuming 100% excretion (EMEA 2008). For effects data, additionally time-weighted average (TWA) PECs
beef cattle, considering low (0.2 mg/kg body wt/d) and high were calculated using FOCUS for 21, 50, and 100 d, resulting
doses of ivermectin (0.5 mg/kg body wt/d) and high (4.4) and in TWA worst-case PECsw of 0.7, 0.37, and 0.22 ng/L,
low (3.6) log KOC values (cf. Table 2), maximum best- and respectively.
worst-case values for PECsw were 1.9 and 29.4 ng/L, EMEA (2008) also suggests that VetCalc could be used
respectively. Likewise, the best- and worst-case PECsediment alternatively to FOCUS (2006). The VetCalc software
was 0.91 and 2.4 mg/kg sediment wet wt, respectively. (Mackay et al. 2005), which was developed specifically for
Although not proposed by EMEA (2008), the experimentally the risk assessment of veterinary pharmaceuticals, offers a
determined total amount of excreted unchanged ivermectin wide range of application forms, animal types, and geographic
(35%) was taken into account as a more realistic approach for and climatic regions, which can be combined for various
the PEC refinement. For the P scenario considering direct scenarios. This results in a large number of potential PECs, so
excretion, this resulted in a worst-case PECsw of 10.3 ng/L we aimed at simulating worst-case conditions with regard to
and a worst-case PECsediment of 0.84 mg/kg sediment wet wt application form, dosage, animal type, and environmental
when using the worst-case assumptions for the applied dose conditions. Single injections to 2-y-old beef (500 kg body wt)
and KOC (Table 11). at 0.5 mg/kg body wt were simulated because they resulted in
For the IR scenario, EMEA (2008) recommends use of the highest PECs and were comparable to FOCUS simulations.
FOCUS (2006) models for refinement of PECs for ground- For worst-case simulations, we used the lowest KOC and
water, surface water, and sediment. The FOCUS ground- highest DT50 in soil (Tables 2 and 3). We did not consider
water model PEARL is required if the concentration of data on degradation in sediment and water and on excretion,
0.1 mg/L is exceeded in the metamodel. (The metamodel is an which are included in the software’s advanced data section,
empirical equation fitted to the outcomes of the PEARL because use of these data is not recommended by EMEA
model and allows for a rough estimation of PECgw as a simple (2008). This resulted in a worst-case estimate of PECsw for
function of KOC and degradation half-life t1/2 in soil.) Because the P scenario of 12.9 ng/L, which is in the same range as the
Table 11. Refined PECs for ivermectin in surface water, sediment, soil, and dung
Maximum PEC
Compartment (scenario) Guidance/model (scenario) Unit Best case Worst case
Surface water PECsw (P) VetCalca ng/L 0.41 12.9

b
PECsw (P; d.e.) EMEA (sorption þ metabolism) ng/L 0.7 10.3
PECsw (IR) FOCUSc (runoff scenarios ng/L 0.77 6.2

R3 and R4-stream)
PECsw (IR) FOCUSd (TWA for 21, ng/L 0.1, 0.07, 0.05 0.70, 0.37, 0.22
50, and 100 d)
PECsw (IR) VetCalca ng/L 0.20 34.7
Sediment PECsed (P; d.e.) EMEAb (sorption þ metabolism) mg/kg wet wt 0.32 (0.83) 0.84 (2.17)
(mg/kg dry wt)f
PECsed (IR) FOCUSc (runoff scenario R3-stream) 0.17 (0.45) 0.25 (0.65)
b
Soil PECsoil (P) EMEA (metabolism) mg/kg dry wt 0.12 0.73
b
PECsoil (IR) EMEA (metabolism) mg/kg dry wt 0.22 2.13
PECsoil (P) VetCalca mg/kg dry wt 1.14 4.80

b g
PECsoil (IR) EMEA (degradation in manure ) mg/kg dry wt 0.44 5.57
b
PECsoil (IR) EMEA (degradation in mg/kg dry wt 0.47 11.4h
manureg and soil)
PECsoil (IR) VetCalca mg/kg dry wt 1.80 10.8
Dung PECdung (P) EMEAb (excretion pattern) mg/kg dung fresh wt 159 (894)e 420 (2365)e
(mg/kg dry wt)
Values in boldface are used for refined risk characterizations. P, IR, and d.e. as described for Table 8.
a
VetCalc software (Mackay et al. 2005).
b
EMEA (2008).
c
FOCUS (2006); maximal annual concentrations for the scenario resulting in the highest value.
d
Maximum time-weighted average (TWA) PECs for 21, 50, and 100 d using FOCUS (2006).
e
Conversion factor dung fresh wt/dry wt ¼ 5.63 (our results).
f
Conversion factor sediment fresh wt/dry wt ¼ 2.6 (EMEA 2008).
g
Since no data for degradation in manure under anaerobic conditions were available, data for degradation in soil–feces mixtures were used (see section
Refinement of PEC estimation).
h
Assuming a scenario of 5 spreading events on grassland with 2-months intervals.
value derived using the EMEA model for the P scenario direct unchanged ivermectin (35%; see above). With this
excretion but is higher than the PECsw (IR) predicted by approach, the highest derived values were the refined PECsoil
FOCUS (Table 11). For best-case simulations, we considered values of 0.73 and 2.13 mg/kg dry wt for the P and IR scenario,
the highest KOC and lowest DT50 in soil as well as the respectively (Table 11).
advanced data for degradation in sediment and water and on For the PECsoil refinement within the IR scenario, EMEA
excretion. This resulted in maximum best-case PECsw of 0.41 (2008) provides a further approach, which considers the
and 0.20 ng/L for the P and IR scenarios, respectively. The DT50 of the pharmaceutical in manure, the storage time of
PEC for groundwater calculated with VetCalc was always manure, and the nitrogen produced during the storage, with
0.000 ng/L during a period of 10 y. It has to be noted that default values given for the latter 2 parameters. Furthermore,
calculations in VetCalc are based on the Leach-P model, it is assumed that the EU nitrogen spreading limit of
which does not simulate particle-bound transport. This means 170 kg N/ha y1 is met by a single spreading event, as is
that the best-case PECsw from VetCalc based on the highest common practice on arable land. Because no data for
KOC may underestimate the concentration in surface waters. degradation in manure under anaerobic conditions were
Although no risk for soil organisms was indicated in phase available, data for degradation in soil–feces mixtures as
II tier A (RQ ¼ 0.48; Table 9), a PECsoil refinement was specified above (see Environmental fate section) were used
performed according to EMEA (2008) and VetCalc (Mackay instead. However, it should be noted that degradation
et al. 2005), taking into account the excretion pattern and the processes in soil–feces mixtures, which normally occur under
degradation potential in manure and soil. A simple model aerobic conditions, might differ significantly from anaerobic
provided by EMEA (2008) estimates the refined PECsoil by degradation processes in manure or slurry. The highest
multiplying the initial PECsoil with the fraction of excreted PECsoil of 11.4 mg/kg dry wt was derived with the worst-case
assumptions of DT50soil/feces ¼ 240 d and DT50soil ¼ 67 d related avermectin B1a (abamectin) in fish was low: bio-
assuming a scenario of 5 spreading events on grassland with 2- concentration factors (BCFs) of 52 and 56 L/kg were obtained
month intervals (Table 11). in two 42-d studies with Lepomis macrochirus (Wislocki et al.
Similar to the refinement recommended in EMEA (2008), 1989; Van den Heuvel et al. 1996). It was hypothesized that
the framework proposed by Montforts (1999) is used in the large molecular size might have led to a reduced
VetCalc to calculate PECsoil. Based on the combinations of membrane permeation and, thus, to a reduced uptake of
pasture usage, manure management, and environmental avermectin B1a by fish.
scenarios described above, VetCalc predicted maximum
PECsoil of 4.80 and 10.8 mg/kg soil dry wt for the worst-case
pasture and intensively reared animals scenario, respectively PHASE II TIER B ENVIRONMENTAL RISK
(Table 11). ASSESSMENT
It should be noted that the worst-case PECsoil derived by
the VetCalc and EMEA models resulted on some occasions in Fate assessment: Semifield level
higher values than the initial PECsoil derived using the total According to VICH (2004), no further fate studies are
residue approach (cf. Table 4). Hence, a risk of ivermectin required for ivermectin in phase II tier B. However, as part of
accumulation in soil over time is demonstrated, probably a semifield study using terrestrial model ecosystems (TMEs),
caused by its apparent slow degradation and high adsorption some information on the actual concentrations of ivermectin
potential. in soil cores were collected (Förster et al. 2010). The TMEs
For the refinement of PECdung, the highest fraction of the were designed and performed as described by Knacker et al.
applied dose excreted in 1 d was considered (EMEA 2008). (2004). Soil cores were collected from a field site near York,
With this fraction (3.31%, see above), the refined maximum United Kingdom, and established in constant environmental
PECdung was 159 and 420 mg/kg dung fresh wt for the lowest chambers. Ivermectin was applied to the surface of the soil
and highest dosage (best and worst case), respectively (Table cores via slurry made from spiked cow dung at 7 different
11). The model provided by EMEA (2008) does not include concentrations (nominal range 0.75–547 mg/kg soil dry wt,
degradation in dung. assuming a soil depth of 1 cm and a density of 1.5 g/cm3).
After destructive sampling on days 7, 28, and 96 following
application, the concentration of the parasiticide was ana-
Outcome of phase II tier A refined ERA lyzed in the uppermost 1 cm of the soil cores. At the highest
Risk quotients were calculated with refined PEC values applied nominal concentration, the ivermectin content in soil
(Table 11) for those species for which a risk had been did not change considerably (36, 27, and 32% of the nominal
indicated based on initial PECs (Table 12). According to the concentration at 7, 28, and 96 d after application, respec-
outcome of the phase II tier A refined risk assessment, further tively), whereas, at the second highest applied concentration
effects testing in phase II tier B is required for aquatic (182 mg/kg dry wt), the measured contents of ivermectin at
crustaceans, fish, sediment-dwelling organisms, and dung the 3 sampling dates were 34, 15, and 21% of the nominal
organisms. Because the log KOC of ivermectin (3.2; Table 1) concentration. Ivermectin concentrations in lower soil layers
is below the trigger value of 4, no potential for bioaccumu- and in lower treatments were below and around the limit of
lation is indicated according to VICH (2004) and, thus, no detection of 0.34 mg/kg dry wt (Pope 2010). Given that
fish bioaccumulation study was performed. This decision was ivermectin was not directly mixed into the soil but was
supported by the fact that bioaccumulation of the closely adsorbed to dung particles applied on the surface of the soil
Table 12. Phase II tier A risk assessment of ivermectin for the most sensitive taxa using refined maximum PECs
Species Unit PNEC PEC RQ
Surface water Daphnia magna ng/L 0.0057 12.9 (P) 2263
34.7 (IR) 6088
Oncorhynchus mykiss ng/L 3.0 12.9 (P) 4.3
34.7 (IR) 11.6
Sediment Daphnia magna mg/kg wet wt 0.0074–0.0012 0.84 (P; d.e.) 114–700
0.25 (IR) 33.8–208
Soil Folsomia candida mg/kg dry wt 30 4.80 (P) 0.16
11.4 (IR) 0.38
Dung Musca autumnalis mg/kg dung fresh wt 0.047 420 (P) 8936
Aphodius constans 1.76 420 (P) 239
Values in boldface indicate a risk. RQ, d.e. as described for Table 8; PNEC ¼ predicted no effect concentration (Tables 8 and 9); PEC ¼ refined predicted
environmental concentration derived for pasture (P) or intensively reared (IR) animals using different models (Table 11).
cores, these data support laboratory results indicating a low models are currently available for addressing this route of
degradation of this compound in soil. exposure. Field studies were performed on 2 farms in order to
The fate of ivermectin was also assessed in an aquatic quantify the potential concentrations of ivermectin entering
semifield mescocosm study (Sanderson et al. 2007). The aquatic systems via runoff. On the first farm, ivermectin was
parasiticide was added to the water column, and concen- applied to cattle as a pour-on treatment on 2 occasions. On
trations in water and sediment were monitored over time. the second farm, ivermectin was given to sheep as an oral
Ivermectin was found to dissipate rapidly from the water drench on 2 occasions. After each treatment, the runoff
column with a dissipation half-life between 3.1 and 5.3 d. behavior of ivermectin was explored over time. Maximum
Dissipation was attributed to partitioning of ivermectin concentrations in runoff following the 2 treatments at the
into sediment and degradation, probably resulting from cattle farm were 85.4 and 4.1 ng/L, whereas at the sheep
photolysis. Analysis of the sediment indicated that, once in farm, maximum runoff concentrations for the 2 treatments
the sediment layer, ivermectin was very persistent, with a were 120.4 and 28.8 ng/L (Sinclair et al. 2008). Using a
half-life of >265 d. proposed factor of 10 for dilution of the runoff in receiving
waters, a maximum surface water concentration of 12 ng/L
arising from runoff from farmyard hard standing areas was
Fate assessment: Field level estimated. This is within the range of worst-case PECs
For the dung compartment, further studies with dung derived for P and IR scenarios using the recommended models
organisms were conducted on the field level (P scenario), (Table 11).
because after PEC refinement in tier A, the risk quotient for
dung fauna was still 1 (Table 12). In these studies, fate of
ivermectin was also investigated, despite the fact that this is Effect assessment: Aquatic and sediment compartment
not explicitly required by VICH (2004). In phase II tier B, 2 D. magna reproduction tests were
Two large field studies were performed within ERAPharm performed within ERAPharm according to OECD 211
in North England and Central Spain, in order to cover the (1998). Because of the analytical limit of quantification for
geographic and climatic diversity of Europe. These studies ivermectin of 1 ng/L, only samples from the highest test
explored exposure by excretion and field degradation of concentration (1 ng/L) of one of the tests were analyzed, in
ivermectin in dung from treated cattle. At various intervals which recoveries of 70 to 120% were measured. Only the
(e.g., 21, 14, 7, 5, 3, 1 d) before placing dung pats on the lowest tested concentration did not cause any effects on D.
pasture, 4 cattle were treated with ivermectin applied magna growth and reproduction, resulting in an LOEC of
subcutaneously at the recommended dose of 0.2 mg/kg body 0.001 ng/L and an NOEC of 0.0003 ng/L (nominal concen-
wt. Six untreated cattle served as the control. Dung pats on the trations; Garric et al. 2007). Thus, acute to chronic ratio
pasture were protected from disturbance by fences and nets. (ACR) for D. magna was 19 000 (Table 13), which suggests
The first study performed near York, United Kingdom, further chronic testing using more realistic exposure con-
focussed on excretion rates and persistence of ivermectin as ditions and additional taxonomic groups.
well as on degradation in dung and movement from dung to Although the risk characterization in tier A indicated a risk
soil. The concentrations found in dung samples and soil below for freshwater fish, no further fish testing was performed,
dung pats are well within the range determined at farm sites considering the much higher sensitivity of daphnids. To our
in England (Boxall et al. 2006). Almost no transport from the knowledge, no long-term effects data for fish after water
dung to the soil (0-1 cm depth) was observed. There was no exposure to ivermectin are available. However, Johnson et al.
apparent degradation of ivermectin (at 1.3 mg/kg dung dry (1993) investigated long-term toxicity of ivermectin to 4 fish
wt) within the duration of the study (38 d), confirming that species after dietary exposure over 50 d. Although the fish
this substance is highly persistent in dung under field species differed in their ability to tolerate ivermectin, no
conditions (Pope 2010). mortality occurred at the lowest dose of 50 mg/kg fish
Similar results were reported from the second field study, administered every other day. These results suggest that, as
which was performed close to Madrid, Spain. Twenty-eight d expected based on the mode of action of ivermectin, fish are
after deposition, a maximum ivermectin concentration of considerably less sensitive than invertebrates.
4.0 mg/kg soil dry wt was found in the uppermost 2 cm below Toxicity tests were performed with the nematode Caeno-
the dung pats, which contained maximum ivermectin rhabditis elegans in water-only and water–sediment test
concentrations of 90 to 110 mg/kg dung fresh wt (correspond- systems according to ISO (2008) and with L. variegatus and
ing to 450-550 mg/kg dung dry wt) and considerably less Chironomus riparius in water–sediment test systems according
(<1.4 mg/kg dry wt) in the layer of 2- to 5-cm depth to OECD 218 (OECD 2004d). For C. elegans, reproduction
(Römbke, Barrett, et al., 2010). These results confirm the was the most sensitive endpoint resulting in NOECs of
conclusions derived from laboratory fate studies. 1.0 mg/L in the water-only and 100 mg/kg sediment dry wt
The present results on slow degradation in dung agree with in the water–sediment test (Table 13). The toxicity test with
previous field studies. Suarez et al. (2003) estimated a DT50 C. riparius was performed using spiked artificial sediment.
of up to 180 d in cattle dung (180 d after deposition of the Urtica powder was added to the sediment before application
dung pats, 10–57% of the initially applied ivermectin of ivermectin; no additional feeding was provided during the
concentration was detected), whereas Sommer et al. (1992) test. An overall NOEC of 3.1 mg/kg sediment dry wt was
observed no biodegradation during 45 d. These data indicate derived, with dry wt (growth) of the larvae as most sensitive
that slow degradation in soil–dung mixtures can be expected. endpoint. In the toxicity test with L. variegatus, spiked
One route of entry of topical ectoparasiticides to the artificial sediment was used for exposure and Urtica and
aquatic environment that is described by EMEA (2008) is cellulose powder as food source. At concentrations 500 mg/
runoff from farmyard hard standing areas. However, no kg sediment dry wt, ivermectin had a significant effect on
Table 13. Phase II tier B aquatic and sediment effect studies
Water Daphnia magna OECD 211 (1998) NOEC21 d, reprod. ¼ 0.0003 ng/L Garric
et al. (2007)
Caenorhabditis ISO/CD 10872 (2008) NOEC96 h, reprod. 1.0 mg/L This study
elegans (nematode) (water-only exposure)
Sediment C. elegans ISO/CD 10872 (2008) NOEC96 h, reprod. ¼ 100 mg/kg dry wt This study
(sediment exposure)
Chironomus riparius OECD 218 (2004d) NOEC10 d, larval growth ¼ 3.1 mg/kg dry wt Egeler
(insect larvae) et al. (2010)
Lumbriculus variegatus OECD 225 (2007) NOEC28 d, reprod., biomass ¼ 160 mg/kg dry wt Egeler
(benthic oligochaete) et al. (2010)
Benthic communities Natural sediments and meiofauna community: Brinke

overlying water (224 d)b, NOEC224 d ¼ 6.2 mg/kg sedim. dry wt et al. (2010)
abundance and community Nematodes community:
composition NOEC224 d ¼ 0.6 mg/kg sediment dry wt
Water–sediment D. magna, C. riparius Two-species study (51 d)b, D. magna: NOECsurvival, biomass ¼ 53 mg/kg Schweitzer
abundance and biomass dung dry wt et al. (2010)
(D. magna), survival, C. riparius:
growth and emergence NOEClarval survival, larval growth, emergence ¼
(C. riparius)c 263 mg/kg dung dry wt
Cladoceran community Aquatic mesocosm NOEC10–97 d, species richness <30 ng/Ld Sanderson
(265 d)b, abundance et al. (2007)
and species richness
a
All effect concentrations refer to nominal concentrations.
b
Test duration.
c
Application of ivermectin with spiked dung.
d
Significant effects were observed at the lowest nominal concentration (30 ng/L). Measured concentrations (d 10–97) were below the detection limit of 1 ng/L.
survival and reproduction and total biomass of L. variegatus application to cattle (Lumaret et al. 2007). For the
(Egeler et al. 2010). chironomids, an overall NOEC of 263 mg ivermectin/kg dung
In addition to the requirements of EMEA (2008), effects of dry wt was derived. With an NOEC of 53 mg ivermectin/kg
the parasiticide on the community level were investigated in dung dry wt, the daphnids were slightly more sensitive (Table
an indoor microcosm using sediment (0.15% TOC) from a 13). At all tested concentrations, ivermectin could not be
freshwater habitat in Germany with indigenous benthic detected in the water phase (limit of quantification 1 ng/L).
communities (Brinke et al. 2010). The sediment was spiked The high toxicity to cladocerans was confirmed in a long-
with 0.6, 6.2, and 31 mg ivermectin/kg dry wt. After 7, 14, 28, term (265 d) aquatic mesocosm study (Sanderson et al.
56, 112, and 224 d of exposure, abundance and composition 2007). At the lowest nominal ivermectin concentration of
of the meiofauna were assessed. The effect of ivermectin on 30 ng/L, cladoceran species richness, the most sensitive
free-living nematodes, as part of the meiofauna community, endpoint of this study, was significantly affected between
was investigated at the species level. Results were analyzed d 10 and 97. Copepod species richness and abundance of
with univariate and multivariate methods, and principle Ephemeroptera were significantly affected at some but not all
response curves were fitted and statistically tested with sampling dates during the study period. Measured ivermectin
Monte Carlo permutation. NOECs of 6.2 and 0.6 mg/kg concentrations in the water phase during this period were
sediment dry wt were derived for the meiofauna and the initially about 6 ng/L but dropped to below the detection
nematode communities, respectively (Table 13). limit (1 ng/L); concentrations in sediment were about 25 ng/
To simulate direct excretion from pasture animals into kg sediment fresh wt. Full recovery of the cladoceran and
surface waters, a 2-species test using a water–sediment test copepod community and of abundance of Ephemeroptera
system was performed, in which ivermectin was applied via was observed during the following spring, on days 229 and
dung, and long-term effects (51 d) on D. magna and C. 265 of the study.
riparius were evaluated (Schweitzer et al. 2010). Chironomid
larvae and daphnids were exposed via cattle dung spiked with
ivermectin (11, 53, 263, and 1314 mg/kg dung dry wt). The Effect assessment: Terrestrial compartment
highest ivermectin concentration corresponds to the typical Although no further guidance on phase II tier B effects
maximum concentration in dung a few days after topical testing with soil arthropods is provided by VICH (2004),
laboratory tests with additional species as well as semifield lans were clearly more sensitive than the predatory mite (data
and field studies are mentioned by EMEA (2008) as possible not shown; to be published elsewhere). In a similar test with
further procedures. Although not required according to the only 2 species, F. fimetaria and H. aculeifer, the EC10 for the
outcome of the ERA performed so far for ivermectin, collembolan was even lower (0.02 mg/kg soil dry wt; Jensen
additional effect studies with soil organisms at the laboratory et al. 2009; Table 14).
and semifield levels were carried out to verify the above- In the 2 field studies performed in York and Madrid, dung
mentioned guidance documents and to make a profound was collected after treatment of cattle with ivermectin (see
assessment of the effects of ivermectin on terrestrial Fate assessment: Field level section). After homogenization,
organisms. standardized dung pats (0.5 kg wet wt, 15 cm diameter) were
At the laboratory level, a chronic test with the predatory placed randomly on the meadow sites. Effects on abundance
mite Hypoaspis aculeifer was performed in artificial soil of dung organisms and soil invertebrates below the dung pats
according to a recently developed guideline (OECD 2008b). as well as dung decomposition were studied for up to 3
After 16 d of exposure, reproduction of the mites was months after the start of the test. In both studies, abundance
affected, with an EC50 of 17.8 mg/kg soil dry wt (Table 14; of dung flies was strongly impacted. The number of dung-
Römbke, Krogh, et al. 2010). inhabiting beetles was initially reduced but reached control
At the semifield scale, 3 methods with different levels of levels again at later sampling dates (Table 14). No effects
complexity were used. Two of them are classified as gnoto- were found on abundance of soil microarthropods, which
biotic, i.e., test systems are prepared using sieved soil and probably is due to the low concentrations of ivermectin found
introduced test organisms: the MS3 multispecies soil system below treated dung pats (<0.001–0.005 mg/kg soil dry wt).
(Boleas et al. 2005) and the SMS soil multispecies system Decomposition of dung pats was affected at the Madrid site at
(Cortet et al. 2006), whereas the terrestrial model ecosystems a level of 780 mg/kg dung dry wt (Römbke, Barrett, et al.,
(TMEs) are undisturbed soil cores, i.e., soil structure as well 2010; cf. Table 9).
as the local soil organism community have not been changed
(Knacker et al. 2004). The MS3 multispecies soil system
combines the toxicity assessment of soil and leachates. The Use of short-term vs. long-term PECs for surface water
leachate toxicity on D. magna was the most sensitive For the P scenario, the refinement of PECsw according to
endpoint (data not shown; to be published elsewhere). the EMEA models considers sorption properties and data on
However, this endpoint cannot be directly incorporated into metabolism of the pharmaceutical (EMEA 2008). The factor
the soil risk assessment and requires a targeted assessment time is not considered in these models, and a basic assumption
(Tarazona et al. 2010). In the TME study, ivermectin was is that the total residue of unchanged parent compound is
applied via slurry to soil cores from a field site near York that excreted within 1 d. However, in the risk characterization of
were kept in constant environmental chambers (see Environ- phase II tier B, the refined PECsw is compared with the PNEC
mental fate section). No effects of ivermectin were found at derived from chronic effects data.
the tested concentrations on soil respiration and the numbers The highest fraction of ivermectin is excreted within the
of nematodes and enchytraeids. The endpoint affected most first days, with a maximum of 3.31% on day 5 after
strongly was the change in the microarthropod community. application to cattle (see Refinement of PEC estimation
Detailed results of this study will be published elsewhere section). With this value for the scenario direct excretion
(Förster et al. 2010). These results confirm that ivermectin is (d.e.), a transient exposure peak for surface waters (short-
affecting arthropods more strongly than other soil organism term PECsw d.e.) can be calculated, when refining the default
groups, as had already been concluded from the laboratory value for the fraction of the total absorbed dose excreted into
tests. the stream (Fe) of 0.01 (EMEA 2008). This refinement results
Results from the SMS test system have not yet been in a best-case and worst-case short-term PECsw d.e. of 0.06
published. However, as in the laboratory tests, all collembo- and 1.0 ng/L, respectively. This short-term PEC could then be
Table 14. Phase II tier B terrestrial effect studies
Soil Hypoaspis aculeifer OECD (2008b) NOECreprod. ¼ 3.2 mg/kg dry wt Römbke, Krogh,
(predatory mite) (artificial soil, EC50reprod. ¼ 17.8 mg/kg dry wt et al. (2010)
TOC 3.6%)
Folsomia fimetaria, 2-species test F. fimetaria: EC10reprod. ¼ 0.02 mg/kg dry wt Jensen et al. (2009)
H. aculeifer system (21 db, H. aculeifer: EC10reprod. ¼ 0.04 mg/kg dry wt
(collembolan, mite) reproduction)
Dung Wildlife communities: Field study Madrid: NOECbeetles ¼ 0.81 mg/kg dung dry wtc Römbke, Barrett,
dung beetles, dung Abundance, dung NOECflies <0.31 mg/kg dung dry wtc et al. (2010)
flies decomposition (86 d)b NOECdecomp. <0.78 mg/kg dung dry wt
a
Effect concentrations refer to nominal concentrations.
b
Test duration.
c
Abundance of beetles and flies was investigated during the first 28 d of the study.
compared with a PNEC derived from acute toxicity data, in assessment for the implementation of REACH (ECHA
the case of ivermectin, the EC50 for D. magna (Table 8). 2008), accumulation in soil and soil-dwelling organisms also
has to be assessed. Therefore, further studies are required for
a reliable PBT assessment of ivermectin, e.g., the determi-
Screening for PBT properties
nation of the KOW (DOW) according to OECD 107 or 117
Persistent, bioaccumulative, and toxic (PBT) as well as very and the assessment of the BCF (BAF) for water, sediment, or
persistent and very bioaccumulative (vPvB) substances are of soil.
particular concern, because their effects are difficult to reverse
and are often not detected at an early stage. Therefore, EMEA
(2008) suggests assessing these substance properties according Risk characterization
to the technical guidance document on ERA of industrial The effects data derived according to and beyond the
chemicals and biocides (EC 2003). According to the data requirements of VICH (2004) and the maximum refined
indicted in Tables 3 and 13, ivermectin fulfills the P criterion PECs (Table 11) were used for risk characterization (Table
(degradation half-life >120 d in freshwater sediment) and the 15). Because long-term effects data are available for at least 3
T criterion (chronic NOEC <0.01 mg/L) as indicated in EC trophic levels within the respective compartments, an AF of
(2003). Concerning the bioconcentration factor (BCF) and, 10 was generally applied to the lowest NOEC values to derive
thus, the B criterion (BCF >2,000), no measured data are the PNEC according to EMEA (2008). An AF of 1000 was
available for ivermectin. By using the simple formula applied to the short-term effects data for D. magna. In this
provided by EMEA (2008), a BCF of 100 is estimated based case, the PNEC was compared with the short-term PEC as
on the KOW. This formula tends to overestimate the BCF for described in the previous section. This risk characterization
substances with a molecular weight above 700 g/mol, but it resulted in a high acute risk indicated for D. magna when
can be used to derive an initial worst-case estimate (EMEA exposed to water concentrations that might occur transiently
2008). Note that the reliability of the available KOW could during the peak excretion of ivermectin by cattle on pasture.
not be checked (see Table 1). Given the dissipation and For the field study, no AF was applied to the NOECs for dung
sorption properties of ivermectin (Tables 2 and 3), it is organisms, because no guidance for this is given by EMEA
assumed that accumulation in sediment and sediment-dwell- (2008).
ing organisms may occur and, hence, that biomagnification The risk characterization using long-term effects data for
processes additionally play a role in the aquatic environment. aquatic and sediment organisms (D. magna and C. riparius) as
It should be noted that, according to the guidance on PBT required according to VICH (2004) resulted in an indication
Table 15. Summary of phase II tier B risk assessment for ivermectin for different compartments
Species or biological Effect

parameter concentrationa Unit AF PNEC PEC RQ
Surface water D. magna 5.7b ng/L 1000 0.0057 1.0 (P; d.e.)c 175
D. magna 0.0003 10 0.00003 12.9 (P) 4.3 T 105
10.3 (P; d.e.) 3.4 T 105
0.70 (IR)d 2.3 T 104
D. magna (2-species) <1 10 <0.1 10.3 (P; d.e.) >103
Sediment C. riparius 3.1 mg/kg sed. 10 0.31 2.17 (P; d.e.) 7

dry wt
0.65 (IR) 2.1
Benthic communities 0.6 0.06 2.17 (P; d.e.) 36
0.65 (IR) 10.8
Soil F. fimetaria (2-species test) 20e mg/kg soil 10 2 4.80 (P), 11.4 (IR) 2.4, 5.7
dry wt
f
Dung Dung fly community (field) <0.31 mg/kg dung <0.31 2.365 (P) >7.6
dry wt
Dung decomposition (field) <0.78 <0.78 >3.0
Values in boldface indicate a risk. AF, PNEC, RQ, and d.e. as described for Table 8; PEC ¼ refined maximum predicted environmental concentration derived for
pasture (P) or intensively reared (IR) animals as shown in Table 11.
a
NOEC values (long-term) as shown in Table 13 and 14; one exception is marked with ‘‘c.’’
b
Short-term EC50 as shown in Table 8.
c
Short-term PECsw d.e. (see section Use of short term vs. long term PECs for surface water).
d
As a more realistic approach, the TWA PECsw for 21 d (Table 11) was used.
e
Refers to EC10.
f
No guidance on assessment factors for such field studies is available. However, even without any AF, a risk is indicated for dung insects.
of risk for these compartments. While the RQ for sediment on pasture. Because the persistency (P) and toxicity (T)
organisms was between 2.1 and 36, the RQ for daphnids was criteria (EC 2003) for ivermectin are fulfilled, and dissipation
>105, indicating a very high risk for aquatic invertebrates. and sorption properties suggest that bioaccumulation and
The aquatic 2-species study using dung spiked with ivermec- biomagnification processes may play a role for ivermectin in
tin also resulted in a risk indication for daphnids. For sediment the aquatic environment, further studies regarding the B
organisms, the risk demonstrated in phase II tier B assessment property are necessary for supporting the PBT assessment of
was confirmed for both the IR and the P scenarios using the ivermectin.
effects data from the study with natural benthic communities Within the present case study, it was not in all cases feasible
(Table 15). to perform the studies requested by VICH (2000, 2004) and
For soil, a risk was now indicated for the P and IR scenarios EMEA (2008): no data were generated regarding octanol–water
based on the effects on collembolans observed in the partitioning, degradation in manure, or effects on fish early life
terrestrial 2-species study. Based on the results of the field stages. These data gaps increase the degree of uncertainty for
studies, mainly the Madrid study, RQs for dung organisms some parts of the ERA. In addition, the literature data had to be
and dung decomposition were above 1 (Table 15), but no risk used in a few cases for which no assessment of reliability was
was indicated for soil invertebrates. Hence, the phase II tier B feasible. For example, the only available measured KOW was
risk assessment for ivermectin indicated risk for the compart- taken from dossier data (USFDA 1990), for which no details on
ments surface water, sediment, soil, and dung. the experimental method are available. Furthermore, some
data used in the ERA were recently submitted for publication
DISCUSSION and are still being reviewed. These facts further contribute to
The ERA of ivermectin, which was performed mainly the uncertainty of the present ERA for ivermectin, which
according to VICH (2000, 2004) and EMEA (2008), initially should therefore partially be regarded as preliminary. However,
resulted in an indication of risk for surface water, sediment, each risk assessment suffers from some degree of uncertainty
and dung (Table 16, phase II-A). For the aquatic compart- regarding the available data, the extrapolation, or the risk
ment, this risk was based mainly on the extremely high characterization. For this reason, assessment factors were
toxicity of ivermectin to daphnids, with long-term effects in employed at all tiers of the ERA.
the low picograms-per-liter range and a PNEC in the The environmental concentrations used in the risk assess-
femtograms-per-liter range. Although a risk was also indi- ment are predicted values using models, several of them
cated for fish in phase II tier A and hence chronic fish testing provided by EMEA (2008), all of which offer a number of
was required in tier B according to EMEA (2008), further choices to the applicant on how to parameterize the models.
phase II tier B studies, such as a fish early life-stage test This leads to a range of PECs resulting in best- and worst-case
(OECD 1992), were not performed, given the much higher risk characterizations but also to a higher level of uncertainty.
sensitivity of daphnids. Thus, the phase II tier B ERA for It would be helpful to have more detailed guidance on how
aquatic species is based on daphnids only. these risk characterizations should be systematically eval-
For sediment, a risk was indicated at all tiers of the ERA uated, reported, and interpreted, and also which scenarios
when using data from standardized single-species toxicity should be chosen (Schneider et al. 2007). This issue is critical
tests with sediment-dwelling organisms and from a mesocosm if RQs are close to 1 and model parameterization, choice of
study with natural benthic communities. Furthermore, and application mode, and exposure scenarios have an important
beyond the assessment according to the VICH and EMEA effect on exposure concentrations.
guidelines, a high acute risk for D. magna was also indicated A limited amount of monitoring data is available with
when comparing the PNEC derived from Daphnia short-term maximum ivermectin concentrations in the water column of
effects data with the short-term PEC that might occur ditches in the pasture environment of <0.2 ng/L (Boxall et al.
transiently during peak excretion of ivermectin by cattle kept 2006). This suggests that concentrations in surface waters
Table 16. Overview of the overall risk assessment for ivermectin according to the tiered approach recommended by VICH (2000, 2004) and
EMEA (2008) and additional studies performed within the present case study
Organism Phase II-A initial PEC Phase II-A refined PEC Phase II-B refined PEC
Surface water Algae No risk Not required Not required
Daphnia Risk Risk Risk
Fish Risk Risk No data available
Sediment Chironomids and Risk Risk Risk

benthic communities
Soil Plants No risk Not required Not required
Earthworms No risk Not required Not required
Collembolans No risk Not required Not required, but

risk in 2-species study
Dung Dung beetles and Risk Risk Risk (field study)

dung flies
might be lower than the models predict. However, high The refined PECsoil values for ivermectin in the IR and P
toxicity of ivermectin to daphnids was observed at concen- scenarios, which integrate information on adsorption, degra-
trations clearly below the detection limit for this compound dation, and excretion, were by factors of 1.9 and 2.3 higher,
in water. respectively, than the initial PECsoil. At phase I, the tiered
Previous ERAs of products containing ivermectin had approach of VICH (2000) and EMEA (2008) does not
revealed no concern for the aquatic compartment (e.g., consider properties of the active ingredient, which might
USFDA 1996, 2001). This was based on the fact that the result in potential for accumulation in soil (at this stage, only
high sorption and low leaching potential of ivermectin had degradation in manure can be considered, as far as such data
suggested little potential of exposure of aquatic species. are available, to reduce the initial PECsoil). In consequence,
However, Garric et al. (2007) showed that extremely low exposure to compounds with high-adsorption and low-
ivermectin concentrations, which can be expected despite the degradation properties can be underestimated using the initial
sorptive properties of the parasiticide, may cause effects on PECsoil.
daphnids. Moreover, the additional aquatic 2-species study Data from the literature (e.g., Madsen et al. 1990; Floate
simulating direct excretion into surface water (Schweitzer 1998a, 1998b; Krüger and Scholtz 1998a, 1998b; Lumaret
et al. 2010) confirmed the risk for daphnids. For the soil and Errouissi 2002) as well as from studies of structural and
compartment, the risk assessment in phase II tier A according functional endpoints within ERAPharm show that higher
to VICH (2004) and EMEA (2008) did not reveal a risk, tier evaluation of effects under field conditions provides
whereas the terrestrial 2-species study, which was performed information essential for the ERA. Despite the large amount
beyond the requirements of the guidelines, indicated a risk for of data, regulatory guidance on how to conduct field studies is
collembolans. not yet available. Nevertheless, it can be concluded that the
In USFDA (2001), transient effects of ivermectin on dung- decomposition of dung is a promising parameter for assessing
insect populations were regarded as not relevant for the the impact of parasiticides on ecosystem function and services
environment, assuming rapid degradation of ivermectin in (Millennium Ecosystem Assessment 2003; Svendsen et al.
sunlight. However, the field studies performed within 2003). In addition, the dominance spectrum or species
ERAPharm showed that ivermectin poses considerable risk number of soil or dung communities might also be relevant
to dung fauna and dung decomposition. The field studies may endpoints. To date, no clear criteria or plausible recommen-
have overestimated the risks for dung organisms, because dations are available for a tiered effects assessment in the dung
farmers are usually only treating animals for a few days each compartment. Because these issues have successfully been
year. Hopefully, in the near future, improved management addressed in aquatic ecotoxicology (see, e.g., Giddings et al.
practice will lead to a more targeted treatment of livestock 2002), it should also be possible to provide suitable guidance
parasitosis and, thus, to a reduction of effects on dung for the terrestrial compartment. Finally, research is needed to
organisms. However, further research is needed to improve check which scale of field studies (in ERAPharm studies up to
the understanding of the interactions among infectious diseases 1 ha) is appropriate, insofar as larger scales probably are
caused by parasites, the life cycle of dung organisms, and the required for studying issues such as the recovery of dung
possible impact of parasiticides on the dung fauna as well as organisms.
livestock management and the veterinary practice to treat such Based on the outcome of the ERA, risk-mitigation
diseases. One possible approach to extrapolate results of measures may be necessary to avoid the possible entry of
laboratory and field studies to the actual agricultural situation ivermectin into the environment. The requirement and
might be to employ population-modeling approaches together definition of risk-mitigation measures within the registration
with information on ivermectin usage, excretion character- and authorization procedures for veterinary pharmaceuticals
istics, and animal husbandry methods as used by Boxall et al. is a common practice (Koschorreck and de Knecht 2004).
(2007). It should be noted that new concepts for higher tier However, different entry pathways resulting from different
dung-fauna studies were discussed recently with dung fauna application methods have to be considered, and measures
experts including long-term laboratory tests with sublethal have to be specifically tailored. Therefore, further research is
endpoints (Adler and Römbke 2008). needed to identify appropriate risk-mitigation measures for
For both the IR and the P scenario, initial PECsoil values for ivermectin containing veterinary medicinal products. It may
ivermectin were below the trigger value (action limit) of be appropriate, for example, to recommend to farmers to
100 mg/kg soil dry wt. According to VICH (2000), the ERA keep treated animals away from watercourses for a certain
of ivermectin-containing products applied exclusively to time following treatment in order to reduce the risk to surface
intensively reared animals stops after phase I, because waters. The time intervals should be fixed based on excretion
concentrations below the trigger value are not expected to data for the treated animal species, drug formulation, and
result in risks for the environment following the IR animal route of application. Mitigation measures may also be
exposure scenario (see also Schmitt et al. 2010). However, for necessary to reduce the risk to dung organisms. The
antiparasitic products intended for animals reared on pasture, practicability and efficacy of potential mitigation approaches
phase II testing is required independently of the predicted remains to be established.
environmental exposure concentration. The risk assessment
presented in this paper clearly demonstrates that, for both the CONCLUSIONS
IR and the P scenario, an unacceptable risk is determined for The results of the present case study clearly demonstrate
all investigated compartments (surface water, sediment, soil, that, with regard to its environmental aspects, ivermectin is a
and dung). Hence, the action limit of 100 mg/kg soil dry wt is substance of high concern. The ERA of ivermectin was
not protective for substances such as ivermectin used on performed mainly according to international and European
intensively reared animals. Possible alternatives to the action guidelines (VICH 2000, 2004; EMEA 2008), using a large
limit are discussed by Schmitt et al. (2010). number of new data on fate and effects of ivermectin and
additional results from 2-species, multispecies, semifield, and Duce IR, Scott RH. 1985. Actions of dihydroavermectin B1a on insect muscle. Br J
field studies obtained within the ERAPharm project. Previous Pharmacol 85:395–401.
ERAs for ivermectin had revealed no concern for the aquatic Dybas RA. 1989. Abamectin use in crop protection. In Campbell WC, editor.
compartment. Effects on dung-insect populations had been Ivermectin and abamectin. New York (NY): Springer-Verlag. p 287–310.
considered as transient and thus not relevant. In contrast to [EC] European Commission. 2003. Technical Guidance Document (TGD) on risk
these ERAs, the present case study—although in part assessment in support of Commission Directive 93/67/EEC on risk assessment
preliminary—clearly demonstrates unacceptable risks (e.g., for new notified substances, Commission Regulation (EC) No 1488/94 on risk
for daphnids and dung organisms) and, hence, suggests the assessment for existing substances, Directive 98/8/EC of the European
necessity of reassessing ivermectin containing veterinary Parliament and of the Council concerning the placing of biocidal products
on the market. Ispra (IT): European Commission Joint Research Centre. EUR
medicinal products. Furthermore, the case study indicates
20418 EN/1-4.
several gaps in the existing guidelines, which should be
[ECHA] European Chemicals Agency. 2008. Guidance on information requirements
considered within guideline revision processes.
and chemical safety assessment. Chapter R.11: PBT assessment.
Acknowledgment—This work was funded by the European
Edwards CA, Atiyeh RM, Römbke J. 2001. Environmental impact of avermectins.
Union under the sixth framework program in the STREP
Rev Environ Contam Toxicol 171:111–137.
‘‘Environmental risk assessment of pharmaceuticals’’ (ERA- Egeler P, Gilberg D, Fink G, Duis K. 2010. Chronic toxicity of ivermectin to the
Pharm; SSPI-CT-2003-511135). We thank Jan Koschorreck benthic invertebrates Chironomus riparius and Lumbriculus variegatus. J Soils
and 3 anonymous reviewers for valuable comments on the Sediment 10:368–376.
manuscript. The views expressed herein do not necessarily [EMEA] European Medicines Agency. 1997. Note for guidance: Environmental risk
reflect the views or policies of the authors’ agencies. assessment for veterinary medicinal products other than GMO-containing and
immunological products. London (UK): EMEA. Final report EMEA/CVMP/055/
REFERENCES 96.
Adler N, Römbke J. 2008. Minutes of the Third Dung Fauna Expert Workshop, 2008 [EMEA] European Medicines Agency. 2008. Revised guideline on environmental
June 25-26; Berlin, Germany. impact assessment for veterinary medicinal products in support of the VICH
Avdeef A, Bendels S, Tsinman KL, Kansy M. 2007. Solubility excipient classification guidelines GL6 and GL38. London (UK): Committee for Medicinal Products for
gradient maps. Pharmacol Res 24:530–545. Veterinary Use (CVMP), EMEA. EMEA/CVMP/ERA/418282/2005-Rev.1.
Blume RR, Younger RL, Aga A, Myers CJ. 1976. Effects of residues of certain Escher BI, Berger C, Bramaz N, Kwon H-W, Richter M, Tsinman O, Avdeef A. 2008.
anthelmintics in bovine manure on Onthophagus gazella, a non-target Membrane-water partitioning, membrane permeability, and baseline toxicity of
organism. Southwest Entomol 2:100–103. the parasiticides ivermectin, albendazole, and morantel. Environ Toxicol Chem
Boelsterli UA. 2003. Mechanistic toxicology—The molecular basis of how 27:909–918.
chemicals disrupt biological targets. London (UK): Taylor & Francis. 312 p. Fernandez C, San Andrés M, Porcel MA, Rodriguez C, Alonso Á, Tarazona JV. 2009.
Boleas S, Alonso C, Pro J, Babı́n MM, Fernández C, Carbonell G, Tarazona JV. 2005. Pharmacokinetic profile of ivermectin in cattle dung excretion, and its
Effects of sulfachlorpyridazine in MS3-arable land: A multispecies soil system associated environmental hazard. Soil Sediment Contam 18:564–575.
for assessing the environmental fate and effects of veterinary medicines. Fink DW. 1988. Ivermectin. In: Florey K, editor. Analytical profiles of drug
Environ Toxicol Chem 24:811–819. substances. New York (NY): Academic. p 155–184.
Boxall ABA, Fogg LA, Baird DJ, Lewis C, Telfer TC, Kolpin D, Gravell A, Pemberton E, Floate KD. 1998a. Off-target effects of ivermectin on insects and on dung
Boucard T. 2006. Targeted monitoring study for veterinary medicines in the degradation in southern Alberta, Canada. B Entomol Res 88:25–35.
environment. Bristol (UK): Environment Agency. Science Report SC030183/SR. Floate KD. 1998b. Does a repellent effect contribute to reduced levels of insect
Boxall ABA, Fogg LA, Blackwell PA, Kay P, Pemberton EJ. 2002. Review of veterinary activity in dung from cattle treated with ivermectin? B Entomol Res 88:291–297.
medicines in the environment. Bristol (UK): Environment Agency. R&D Technical Floate KD, Wardhaugh KG, Boxall AB, Sherratt TN. 2005. Fecal residues of veterinary
Report P6-012/8TR. parasiticides: Nontarget effects in the pasture environment. Annu Rev Entomol
Boxall ABA, Fogg LA, Blackwell PA, Kay P, Pemberton EJ, Croxford A. 2004. 50:153–179.
Veterinary medicines in the environment. Rev Environ Contam Toxicol [FOCUS] Forum for the Co-ordination of pesticide fate models and their Use. 2006.
180:1–91. Guidance document on estimating persistence and degradation kinetics from
Boxall ABA, Sherratt TN, Pudner V, Pope LJ. 2007. A screening level index for environmental fate studies on pesticides in EU registration. Report of the
assessing the impacts of veterinary medicines on dung flies. Environ Sci Technol FOCUS Work Group on Degradation Kinetics. EC document reference
41:2630–2635. Sanco/10058/2005 version 2.0.
Brinke M, Höss S, Fink G, Ternes TA, Heininger P, Traunspurger W. 2010. Assessing Forbes AB. 1993. A review of regional and temporal use of avermectins in cattle and
effects of the pharmaceutical ivermectin on meiobenthic communities using horses worldwide. Vet Parasitol 48:19–28.
freshwater microcosms. Aquat Toxicol DOI: 10.1016/j.aquatox.2010.04.008. Förster B, Boxall A, Coors A, Jensen J, Liebig M, Pope L, Moser T, Römbke J. 2010.
Campbell WC, Fisher MH, Stapley EO, Albers-Schonberg G, Jacob TA. 1983. Fate and effects of ivermectin on soil invertebrates in terrestrial model
Ivermectin: A potent new antiparasitic agent. Science 221:823–828. ecosystems. Ecotoxicology (submitted).
Chiu S-HL, Sestokas E, Taub R, Buhs RP, Green M, Sestokas R, Vandenheuvel WJA, Garric J, Vollat B, Duis K, Péry A, Junker T, Ramil M, Fink G, Ternes TA. 2007. Effects
Arison BH, Jacob TA. 1986. Metabolic disposition of ivermectin in tissues of of the parasiticide ivermectin on the cladoceran Daphnia magna and the green
cattle, sheep, and rats. Drug Metab Dispos 14:590–600. alga Pseudokirchneriella subcapitata. Chemosphere 69:903–910.
Chiu S-HL, Sestokas E, Taub R, Green ML, Baylis FP, Jacob TA, Lu AYH. 1990. Giddings JM, Brock TCM, Heger W, Heimbach F, Maund SJ, Norman SM, Ratte HT,
Metabolic disposition of ivermectin in swine. J Agric Food Chem 38:2079– Schäfers C, Streloke M. 2002. CLASSIC. Community-level aquatic system
2085. studies-interpretation criteria. Brussels (BE): SETAC-Europe. 43 p.
Cook DF, Dadour IR, Ali DN. 1996. Effect of diet on the excretion profile of Grant A, Briggs AD. 1998. Use of ivermectin in marine fish farms: some concerns.
ivermectin in cattle faeces. Int J Parasitol 26:291–295. Mar Pollut Bull 36:566–568.
Cortet J, Joffre R, Elmholt S, Coeurdassier M, Scheifler R, Krogh PH. 2006. Grønvold J, Stendal Svendsen T, Kraglund H-O, Bresciani J, Monrad J. 2004. Effect of
Interspecific relationships among soil invertebrates influence pollutant the antiparasitic drugs fenbendazole and ivermectin on the soil nematode
effects of phenanthrene. Environ Toxicol Chem 25:120–127. Pristionchus maupasi. Vet Parasitol 124:91–99.
Davies IM, McHenery JG, Rae GH. 1997. Environmental risk from dissolved Gunn A, Sadd JW. 1994. The effect of ivermectin on the survival, behaviour and
ivermectin to marine organisms. Aquaculture 158:263–275. cocoon production of the earthworm Eisenia fetida. Pedobiologia 38:327–333.
Halley BA, Jacob TA, Lu AYH. 1989a. The environmental impact of the use of Krüger K, Scholtz CH. 1998a. Changes in the structure of dung insect communities
ivermectin: environmental effects and fate. Chemosphere 18:1543–1563. after ivermectin usage in a grassland ecosystem. I. Impact of ivermectin under
Halley BA, Narasimhan NI, Venkataraman K, Taub R, Green ML, Erwin N, Andrew W, drought conditions. Acta Oecol 19:425–438.
Wislocki PG. 1992. Ivermectin and abamectin metabolism, differences and Krüger K, Scholtz CH. 1998b. Changes in the structure of dung insect communities
similarities. In: Hutson DH, Hawkins DR, Paulson GD, Struble CB, editors. after ivermectin usage in a grassland ecosystem. II. Impact of ivermectin under
Xenobiotics and food-producing animals. Washington DC. ACS Symposium high rainfall conditions. Acta Oecol 19:439–451.
Series, Vol 503. p 203–216. Lindley D. 1987. Merck’s new drug free to WHO for river blindness programme.
Halley BA, Nessel RJ, Lu AYH. 1989b. Environmental aspects of ivermectin usage in Nature 329:752.
livestock: General considerations. In: Campbell WC, editor. Ivermectin and Löffler D, Römbke J, Meller M, Ternes T. 2005. Environmental fate of
abamectin. New York (NY): Springer-Verlag. p 162–172. pharmaceuticals in water/sediment systems. Environ Sci Technol 39:5209–
Halling-Sørensen B, Nors Nielsen, Lanzky S, Ingerslev PF, Holten F, Lützhøft HC, 5218.
Jørgensen SE. 1998. Occurrence, fate and effects of pharmaceutical substances Lumaret J-P, Alvinerie M, Hempel H, Schallnass H-J, Claret D, Römbke J. 2007. New
in the environment—A review. Chemosphere 36:357–393. screening test to predict the potential impact of ivermectin-contaminated
Hempel H, Scheffczyk A, Schallnaß H-J, Lumaret J-P, Alvinerie M, Römbke J. 2006. cattle dung on dung beetles. Vet Res 38:15–24.
Effects of four veterinary pharmaceuticals on the dung beetle Aphodius Lumaret J-P, Errouissi F. 2002. Use of anthelminthics in herbivores and evaluation of
constans in the laboratory. Environ Toxicol Chem 25:3155–3163. risks for the non target fauna of pastures. Vet Res 33:547–562.
[ISO] International Organization for Standardization. 1999. Soil quality—Inhibition Mackay N, Mason P, Di Guardo A. 2005. VetCalc exposure modelling tool for
of reproduction of collembola (Folsomia candida) by soil pollutants. Geneva veterinary medicines. London (UK): Defra Research Project VM02133.
(CH): ISO 11267. Madsen M, Overgaard Nielsen B, Holter P, Pedersen OC, Brøchner Jespersen J, Vagn
[ISO] International Organization for Standardization. 2004. Soil quality—Effects of Jensen KM, Nansen P, Grønvold J. 1990. Treating cattle with ivermectin: effects
pollutants on Enchytraeidae (Enchytraeus sp.). Determination of effects on on the fauna and decomposition of dung pats. J Appl Ecol 27:1–15.
reproduction and survival. Geneva (CH): ISO 16387. Millennium Ecosystem Assessment. 2003. Ecosystems and human wellbeing: A
[ISO] International Organization for Standardization. 2008. Water quality— framework for assessment. Washington DC. Island. 245 p.
Determination of the toxic effect of sediment and soil samples on growth, Montforts MHMM. 1999. Environmental risk assessment for veterinary medicinal
fertility and reproduction of Caenorhabditis elegans (Nematoda). Geneva (CH): products. Part 1: Other than GMO-containing and immunological products.
ISO/CD 10872. Bilthoven (NL): RIVM. First update report 601300 001.
Jensen J, Diao X, Hansen AD. 2009. Single- and two-species tests to study effects of [NRA] National Registration Authority. 1998. Report on the special review of
the anthelmintics ivermectin and morantel, and the coccidiostatic monensin on macrocyclic lactones. Canberra (AU): Chemical Review Section, NRA. Special
soil invertebrates. Environ Toxicol Chem 28:316–323. Review Series 98.3.
Jensen J, Krogh PH, Sverdrup LE. 2003. Effects of the antibacterial agents tiamulin, [OECD] Organisation for Economic Co-Operation and Development. 1984.
olanquindox and metronidazole and the anthelmintic ivermectin on the soil Earthworm acute toxicity test. Paris (FR): OECD. Guideline for testing of
invertebrate species Folsomia fimetaria (Collembola) and Enchytraeus crypticus chemicals 207.
(Enchytraeidae). Chemosphere 50:437–443. [OECD] Organisation for Economic Co-Operation and Development. 1992. Fish,
Johnson SC, Kent ML, Whitaker DJ, Margolis L. 1993. Toxicity and pathological early-life stage toxicity test. Paris (FR): OECD. Guideline for testing of chemicals
effects of orally administered ivermectin in Atlantic, chinook, and coho salmon 210.
and steelhead trout. Dis Aquat Org 17:107–112. [OECD] Organisation for Economic Co-Operation and Development. 1998.
Kilmartin J, Cazabon D, Smith P. 1996. Investigations of the toxicity of ivermectin for Daphnia magna reproduction test. Paris (FR): OECD. Guideline for the
salmonids. B Eur Assoc Fish Pat 17:58–61. testing of chemicals 211.
Klimisch H-J, Andreae M, Tillmann U. 1997. A systematic approach for evaluating [OECD] Organisation for Economic Co-Operation and Development. 2000.
the quality of experimental toxicological and ecotoxicological data. Regul Adsorption-desorption using a batch equilibrium method. Paris (FR): OECD.
Toxicol Pharm 25:1–5. Guideline for the testing of chemicals 106.
Knacker T, van Gestel CAM, Jones SE, Soares AMVM, Schallnass H-J, Förster B, [OECD] Organisation for Economic Co-Operation and Development. 2002a.
Edwards CA. 2004. Ring-testing and field-validation of a terrestrial model Aerobic and anaerobic transformation in soil. Paris (FR): OECD. Guideline for
ecosystem (TME)—An instrument for testing potentially harmful substances: testing of chemicals 307.
Conceptual approach and study design. Ecotoxicology 13:5–23. [OECD] Organisation for Economic Co-Operation and Development. 2002b.
Kolar L, Kožuh Eržen N. 2006. Veterinary parasiticides—Are they posing an Aerobic and anaerobic transformation in aquatic sediment systems. Paris
environmental risk? Sloven Vet Res 43:85–96. (FR): OECD. Guideline for the testing of chemicals 308.
Koschorreck J, Apel P. 2006. A brief overview on the legal background and the [OECD] Organisation for Economic Co-Operation and Development. 2002c.
regulatory instruments of the environmental risk assessment for Freshwater algae and cyanobacteria, growth inhibition test. Paris (FR):
pharmaceuticals in the EU, USA, Japan, Australia and Canada. In: Ternes TA, OECD. Guideline for the testing of chemicals, draft revised guideline 201.
Joss A, editors. Human pharmaceuticals, hormones and fragrances. The challenge [OECD] Organisation for Economic Co-Operation and Development. 2004a.
of micropollutants in urban water management. London (UK). IWA. p 107–119. Daphnia sp., acute immobilisation test. Paris (FR): OECD. Guideline for
Koschorreck J, de Knecht J. 2004. Environmental risk assessment of testing of chemicals 202.
pharmaceuticals in the EU - A regulatory perspective. In Kümmerer K, [OECD] Organisation for Economic Co-Operation and Development. 2004b.
editor. Pharmaceuticals in the environment. Sources, fate, effects and risks. Enchytraeidae reproduction test. Paris (FR): OECD. Guideline for testing of
Berlin (DE): Springer. p 289–310. chemicals 220.
Kövecses J, Marcogliese DJ. 2005. Avermectins: Potential environmental risks and [OECD] Organisation for Economic Co-Operation and Development. 2004c.
impacts on freshwater ecosystems in Quebec. Montreal (QC): Environment Earthworm reproduction test. Paris (FR): OECD. Guideline for testing of
Canada—Quebec Region, St. Lawrence Centre. Scientific and technical report chemicals 222.
ST-233E. [OECD] Organisation for Economic Co-Operation and Development. 2004d.
Krogh KA, Jensen GG, Schneider MK, Fenner K, Halling-Sørensen B. 2009. Analysis Sediment-water chironomid toxicity test using spiked sediment. Paris (FR):
of the dissipation kinetics of ivermectin at different temperatures and in four OECD. Guideline for the testing of chemicals 218.
different soils. Chemosphere 75:1097–1104. [OECD] Organisation for Economic Co-Operation and Development. 2007.
Krogh KA, Søeborg T, Brodin B, Halling-Sørensen B. 2008. Sorption and mobility of Sediment–water Lumbriculus toxicity test using spiked sediment. Paris (FR):
ivermectin in different soils. J Environ Qual 37:2202–2211. OECD. Guideline for the testing of chemicals 225.
[OECD] Organisation for Economic Co-Operation and Development. 2008a. Shoop W, Soll M. 2002. Ivermectin, abamectin and eprinomectin. In: Vercruysse J,
Determination of developmental toxicity of a test chemical to dipteran dung Rew RS, editors. Macrocyclic lactones in antiparasitic therapy. Oxon (UK): Cabi.
flies (Scathophaga stercoraria L. [Scatophagidae], Musca autumnalis De Geer p 1–29.
[Muscidae]). Paris (FR): OECD. Guideline for the testing of chemicals 228. Sinclair CJ, Ramwell CT, Turnbull G, Bryning G, Lynn R, Franey-Gardner M, Pepper T,
[OECD] Organisation for Economic Co-Operation and Development. 2008b. Fogg L, Boxall ABA. 2008. Assessment and management of inputs of veterinary
Predatory mite (Hypoaspis [Geolaelaps] aculeifer) reproduction test in soil. medicines from the farmyard. Final report to Defra (UK). VM02501.
Paris (FR): OECD. Guideline for the testing of chemicals 226. Sommer C, Steffansen B, Overgaard Nielsen, Gronvold B, Vagn J, Jensen KM,
[OECD] Organisation for Economic Co-Operation and Development. 2009. Brochner Jespersen J, Springborg J, Nansen P. 1992. Ivermectin excreted in
Guidance document on the determination of the toxicity of a test chemical cattle dung after subcutaneous injection or pour-on treatment: Concentrations
to the dung beetle Aphodius constans. Paris (FR): OECD Environmental Health and impact on dung fauna. B Entomol Res 82:257–264.
and Safety Publications. Series on testing and assessment (draft). Steel JW, Wardhaugh KG. 2002. Ecological impact of macrocyclic lactones on dung
Õmura S. 2002. Mode of action of avermectin. In: Õmura S, editor. Macrolide fauna. In: Vercruysse J, Rew RS, editors. Macrocyclic lactones in antiparasitic
antibiotics: Chemistry, biology and practice. New York (NY): Academic. p 571–576. therapy. Oxon (UK): CABI. p 141–162.
Õmura S. 2008. Ivermectin: 25 years and still going strong. Int J Antimicrob Agric Strong L, Brown TA. 1987. Avermectins in insect control and biology: A review. B
31:91–98. Entomol Res 77:357–389.
Oppel J, Broll G, Löffler D, Meller M, Römbke J, Ternes T. 2004. Leaching behaviour Strong L, James S. 1993. Some effects of ivermectin on the yellow dung fly.
of pharmaceuticals in soil-testing systems: a part of an environmental risk Scathophaga stercoraria. Vet Parasitol 48:181–191.
assessment for groundwater protection. Sci Total Environ 328:265–272. Suarez VH, Lifschitz AL, Sallovitz JM, Lanusse CE. 2003. Effects of ivermectin and
Pope L. 2010. Fate and effects of parasiticides in the pasture environment [PhD doramectin faecal residues on the invertebrate colonization of cattle dung. J
thesis]. York (UK): Univ of York. Appl Entomol 127:481–488.
Prasse C, Löffler D, Ternes TA. 2009. Environmental fate of the anthelmintic Svendsen TS, Grønvold J, Holter P, Sommer C. 2003. Field effects of ivermectin and
ivermectin in an aerobic sediment/water system. Chemosphere 77:1321– fenbendazole on earthworm populations and the disappearance of dung pats
1325. from bolus-treated cattle. Appl Soil Ecol 24:207–218.
Rohrer SP, Arena JP. 1995. Ivermectin interactions with invertebrate ion channels. In: Tarazona JV, Escher BI, Giltrow E, Sumpter J, Knacker T. 2010. Targeting the
Marshall Clark J, editor. Molecular action of insecticides on ion channels. environmental risk assessment of pharmaceuticals: facts and fantasies. Integ
Washington DC. American Chemical Society. p 264–283. Environ Assess Manag 6:603–613 (this issue).
Römbke J, Floate KD, Jochmann R, Schäfer MA, Puniamoorthy N, Knäbe S, Lehmhus [USEPA] US Environmental Protection Agency. 1975. Committee on methods for
J, Rosenkranz B, Scheffczyk A, Schmidt T, Sharples A, Blanckenhorn WU. 2009. toxicity tests with aquatic organisms. In Methods for acute toxicity tests with
Lethal and sublethal toxic effects of a test chemical (ivermectin) on the yellow fish, macroinvertebrates and amphibians. Washington DC. USEPA. EPA/660/3-
dung fly Scathophaga stercoraria based on a standardized international ring 75-009.
test. Environ Toxicol Chem 28:2117–2124. [USFDA] US Food and Drug Administration. 1986. Animal drugs, feeds, and related
Römbke J, Alonso Á, Förster B, Fernandez Torija, Jensen C, Lumaret J, Porcel J-P, Cots products; ivermectin injection. Department of Health and Human Services. Fed
MA, Liebig M. 2010. Effects of the parasiticide ivermectin on the structure and Reg 51:27020–27021.
function of dung and soil invertebrate communities in the field (Madrid, Spain). [USFDA] US Food and Drug Administration. 1990. IVOMEC1 (ivermectin) pour-on
Appl Soil Ecol DOI: 10.1016/j.apsoil.2010.05.004. for cattle. Environmental assessment. NADA 140-841 [cited 2010 May 3].
Römbke J, Barrett K, Blanckenhorn WU, Hargreaves T, Kadiri N, Knäbe S, Lehmhus J, Available from: www.fda.gov/downloads/AnimalVeterinary/Development
Lumaret J-P, Rosenkranz B, Scheffczyk A, Sekine T. 2010. Results of an ApprovalProcess/EnvironmentalAssessments/UCM072241.pdf
international ring test with the dung fly Musca autumnalis in support of a [USFDA] US Food and Drug Administration. 1996. IVOMEC1 SR bolus for
new OECD test guideline. Sci Total Environ (accepted article). DOI: 10.1016/ cattle. Environmental assessment. NADA 140-988 [cited 2010 May 3].
j.scitotenv.2010.05.027 Available from: www.fda.gov/downloads/AnimalVeterinary/Development
Römbke J, Krogh KA, Moser T, Scheffczyk A, Liebig M. 2010. Effects of the ApprovalProcess/EnvironmentalAssessments/UCM072316.pdf
veterinary pharmaceutical ivermectin on soil invertebrates in laboratory tests. [USFDA] US Food and Drug Administration. 2001. Updated environmental
Arch Environ Contam Toxicol 58:332–340. assessment for Merial’s avermectin-based products for cattle. [cited 2008
Römbke J, Moser T. 1999. Organisation and performance of an international October 7]. Available from: www.fda.gov/ohrms/dockets/98fr/140-841-
ringtest for the validation of the enchytraeid reproduction test. Berlin (DE): ea00001.pdf
Umweltbundesamt. UBA-Texte 4/99, Vol I and II. Van den Heuvel WJA, Forbis AD, Halley BA, Ku CC. 1996. Bioconcentration and
Sanderson H, Laird B, Pope L, Brain R, Wilson C, Johnson D, Bryning G, Peregrine AS, depuration of avermectin B1a in the bluegill sunfish. Environ Toxicol Chem
Boxall A, Solomon K. 2007. Assessment of the environmental fate and effects 15:2263–2266.
of ivermectin in aquatic mesocosms. Aquat Toxicol 85:229–240. [VICH] International Cooperation on Harmonisation of Technical Requirements for
Schmitt H, Boucard T, Garric J, Jensen J, Parrott J, Péry A, Römbke J, Registration of Veterinary Medicinal Products. 2000. Environmental impact
Straub JO, Hutchinson TH, Sánchez-Argüello P, Wennmalm Å, Duis K. 2010. assessment (EIAs) for veterinary medicinal products (VMPs)—Phase I. VICH GL
Recommendations on the environmental risk assessment of pharmaceuticals— 6, Ecotoxicity Phase I.
Effect characterization. Integ Environ Assess Manag 6:588–602 (this issue). [VICH] International Cooperation on Harmonisation of Technical Requirements for
Schneider MK, Stamm C, Fenner K. 2007. Selecting scenarios to assess exposure of Registration of Veterinary Medicinal Products. 2004. Environmental Impact
surface waters to veterinary medicines in Europe. Environ Sci Technol 41:4669– Assessment for Veterinary Medicinal Products Phase II Guidance. VICH GL 38,
4676. Ecotoxicity Phase II.
Schweitzer N, Fink G, Ternes TA, Duis K. 2010. Effects of ivermectin-spiked cattle Wislocki PG, Grosso LS, Dybas RA. 1989. Environmental aspects of abamectin use in
dung on a water-sediment system with the aquatic invertebrates Daphnia crop protection. In: Campbell WC, editor. Ivermectin and abamectin. New York
magna and Chironomus riparius. Aquat Toxicol 97:304–313. (NY): Springer. p 182–200.

Integr Envir Assess Manag - 2010 - Liebig - Environmental Risk Assessment of Ivermectin A Case Study

Uploaded by

Copyright:

Available Formats

Integr Envir Assess Manag - 2010 - Liebig - Environmental Risk Assessment of Ivermectin A Case Study

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Integr Envir Assess Manag - 2010 - Liebig - Environmental Risk Assessment of Ivermectin A Case Study

Uploaded by

Copyright:

Available Formats

Integrated Environmental Assessment and Management — Volume 6, Supplement 1—pp.

Environmental Risk Assessment of Ivermectin: A Case Study

INTRODUCTION bacteria or invertebrates (Boxall et al. 2004). In this respect

Water solubility (mg/L) 4.0d (m), 4.1e (m), 2.0f (m)

PHASE II TIER A The limited mobility of ivermectin in soils justifies the

Madrid, Ea 8.7 0.0077 0.0045 57 28–56 1.28 104 4.1

Fulton, USAb 6.3 0.025 0.0145 227c n. d. 1.57 104 4.2

Table 3. Transformation of ivermectin in soils and aquatic sediments

Type of study Value Reference

Transformation in soil (OECD 307)a

Dissipation (Madrid soil) DT50 16 d Krogh et al. (2009)

DT90 54 d Krogh et al. (2009)

Dissipation (York soil) DT50 67 d Krogh et al. (2009)

Dissipation (artificial soil) DT50 458 db Krogh et al. (2009)

Transformation in water–sediment systems (OECD 308)

Dissipation: DT50 (water) <0.25 d Prasse et al. (2009)

Dissipation: DT50 (whole system) 127 d Prasse et al. (2009)

Degradation: t1/2 (water) 30 d Our calculations based

Degradation: t1/2 (sediment) 130 d Our calculations based

Degradation: t1/2 (whole system) 87 d Our calculations based

Dissipation: DT50 (water) 2.9 d Löffler et al. (2005)

Dissipation: DT50 (whole system) 15 d Löffler et al. (2005)

Compartment Unit PEC Remark

PECsoil plateau (P)d mg/kg dry wt 0.86/2.14c Beef cattle

PECgw initial (IR) ng/L 3.3–21.5e Weaner pig (PECgw ¼ PECporewater)

PECgw initial (P) ng/L 0.5–7.4e Beef cattle (PECgw ¼ PECporewater)

Table 5. Phase II tier A aquatic short-term effect studies

Test organism Test method Effect concentration Reference

LOEC72 h, yield, growth rate ¼ 1.25 mg/La

USEPA 660/3-75-009 (1975) LC5048 h ¼ 25 ng/La Halley, Jacob, et al. (1989)

Table 6. Phase II tier A terrestrial effect studies with soil organisms

Test organism Test method Effect concentrationa Reference

NOEC56 d, reprod. ¼ 2.5 mg/kg dry wt

EC5056 d, reprod. ¼ 5.3 mg/kg dry wt

LC5028 d ¼ 315 mg/kg dry wt

LC5014 d ¼ 15.8 mg/kg dry wt

EC5028 d, reprod. ¼ 36 mg/kg dry wt

LC5028 d >300 mg/kg dry wt

EC5028 d, reprod. ¼ 1.7 mg/kg dry wt

EC5028 d, reprod. ¼ 1.7 mg/kg dry wt

LC5028 d ¼ 8.4 mg/kg dry wt

Table 7. Phase II tier A terrestrial effect studies with dung organisms

Test organism Test method Effect concentrationa Reference

NOEC28 d, development time ¼ 0.84 mg/kg dung fresh wt

EC503–4 w., emergence ¼ 1.0 mg/kg dung fresh wt

LC5021 d ¼ 880 mg/kg dung dry wt

NOEC21 d, larval survival ¼ 320 mg/kg dung dry wt

LC5021 d ¼ 590 mg/kg dung dry wt

14561/91754 (P; d.e.)

27.7/174 (P; d.e.)

0.33/2.09 (P) mg/kg 0.003/0.02 (P)

Eisenia fetida NOECreprod. ¼ 2.5 mg/kg 10 250 mg/kg 0.003/0.02 (IR)

Folsomia candida NOECreprod. ¼ 0.3 mg/kg 10 30 mg/kg 0.02/0.20 (IR)

0.0057 ng/L 0.0074/0.0012mg/kg dry wt 0.02/9.25 (IR) 2.7/7708 (IR)

0.03/3.18 (P) 4.1/2650 (P)

16.7/675 (P; d.e.) 2257/562 500 (P; d.e.)

Surface water PECsw (P) VetCalca ng/L 0.41 12.9

PECsw (IR) FOCUSc (runoff scenarios ng/L 0.77 6.2

PECsw (IR) VetCalca ng/L 0.20 34.7