Effects of a bioassay-derived ivermectin lowest ... - Springer Link

Ecotoxicology (2013) 22:148–155 DOI 10.1007/s10646-012-1011-3

Effects of a bioassay-derived ivermectin lowest observed effect concentration on life-cycle traits of the nematode Caenorhabditis elegans Marvin Brinke • Peter Heininger Walter Traunspurger

•

Accepted: 23 October 2012 / Published online: 17 November 2012 Ó Springer Science+Business Media New York 2012

Abstract The pharmaceutical ivermectin is used to treat parasitic infections, such as those caused by nematodes. While several studies have demonstrated the severe effects of ivermectin on non-target organisms, little is known about the drug’s impact on free-living nematodes. In the present work, a full life-cycle experiment was conducted to estimate how an ivermectin lowest observed effect concentration derived from a Caenorhabditis elegans bioassay (endpoint reproduction) might translate into effects at the population level of this free-living nematode. The results showed that fecundity decreased to levels similar to those determined in the bioassay after a time of corresponding duration (18.6 % inhibition compared to the control), but the impact then rather weakened until the end of the experiment, at which point the net reproductive rate (R0) was still, but not significantly, reduced by 12.4 %. Moreover, the average lifespan, length of the reproductive period, maximum daily reproduction rate, and intrinsic rate of increase (rm) were significantly reduced by 30.0, 25.9, 11.2, and 3.5 %, respectively. The experiment revealed that a 4-day bioassay is protective enough for C. elegans with

M. Brinke W. Traunspurger (&) Department of Animal Ecology, University of Bielefeld, Morgenbreede 45, 33615 Bielefeld, Germany e-mail: [email protected] P. Heininger Federal Institute of Hydrology (BfG), Am Mainzer Tor 1, 56068 Koblenz, Germany e-mail: [email protected] Present Address: M. Brinke Federal Institute of Hydrology (BfG), Am Mainzer Tor 1, 56068 Koblenz, Germany e-mail: [email protected]; [email protected]

123

respect to ivermectin’s effects on fecundity. However, the pronounced effects of a low drug concentration on survival, a highly elastic trait, may better account for the observed population-level response, i.e., a decrease of rm, than the effects on fecundity. These results emphasize that full lifecycle experiments are valuable for assessment of pollutants, because the effects on several life-cycle traits can be simultaneously measured and integrated into an ecologically relevant parameter, the population growth rate, that reflects a population’s response to a specific pollutant. Keywords Nematodes Ivermectin Life history Life-cycle traits Population Population growth rate

Introduction Ivermectin is a pharmaceutical that is frequently used to treat livestock and pets against endo- and ectoparasitic infections, such as by nematodes or mites, but it is also used in the treatment of some human diseases, such as river blindness (onchocerciasis), which is also caused by a parasitic nematode (Omura 2008). The drug is a semisynthetic derivate of avermectin B1, a macrocyclic lactone originally isolated from the soil actinomycete Streptomyces avermitilis. It consists of at least 80 % 22,23-dihydroavermectin B1a and not more than 20 % 22,23-dihydroavermectin B1b. In terms of environmental risk assessment (ERA), this very potent antiparasitic drug has raised concerns because of its high toxicities for non-target invertebrates (Campbell et al. 1983; Edwards et al. 2001; Liebig et al. 2010). However, little is known about the risk posed by ivermectin to free-living nematodes, which may be particularly vulnerable due to their close phylogenetic relationship to parasitic nematode species. Free-living

Effects of a bioassay-derived ivermectin on C. elegans

nematodes occur in great diversity and at high densities in every type of sediment or soil and they are often the dominant taxon in these habitats that occupies key positions in benthic and soil food webs because nematodes comprise various feeding types (Bongers and Ferris 1999; Traunspurger et al. 2006; Yeates et al. 1993). Thus far, studies of ivermectin’s effects on free-living nematodes, such as Caenorhabditis elegans, have mainly focused on elucidation of the drug’s mode of action, its efficacy, or the mechanisms of resistance to the drug (Ardelli et al. 2009; Arena et al. 1995; Bernt et al. 1998). In addition, ivermectin’s effects on nematode somatic growth and reproduction have been investigated using standard bioassays (Brinke et al. 2011; Liebig et al. 2010). There have also been a few studies in the field (Yeates et al. 2007a, b, 2003) as well as in model ecosystems (Brinke et al. 2010) aimed at determining ivermectin’s effects on nematode communities. However, studies that are estimating the drug’s impact on free-living nematode populations are missing, although this information is valuable for ERAs in order to reduce uncertainty in extrapolating from ecotoxicological effects to ecologically relevant impacts (Forbes et al. 2008). While ivermectin’s effects on specific life-cycle traits, such as survival or fecundity, can be measured in bioassays carried out at the individual level (evaluating mortality or reproduction), an appreciation of the population-wide impact depends on an understanding of the relationship between these traits and whether the effects on one trait might be compensated by other traits or effects (Ferson et al. 1996). Effects on life-cycle traits can be integrated into a single measure, the population growth rate, in which rm = the intrinsic rate of increase or k = finite rate of increase/ population growth factor (=er). This measure is regarded as an ecologically relevant parameter in assessments of a population’s response to pollutants (Forbes and Calow 1999; Van Leeuwen et al. 1985). Furthermore, traits contributing to the population growth rate have different elasticities, meaning that changes in them lead to different degrees of impact on the population growth rate. Thus, a trait with a low elasticity may be sensitive but it does not strongly impact the growth rate; conversely, a less sensitive trait, but one with a high elasticity, may have a high impact on the growth rate (Forbes et al. 2010). Consequently, if, for example, a single trait with low elasticity is sensitive on an individual level in a bioassay, it cannot be assumed to also have a high impact at the population level. Indeed, conclusions drawn from such a bioassay with respect to an ERA would likely be rather overprotective regarding the population level (Forbes et al. 2001). The challenge has been to clearly and consistently identify those single traits that are the most relevant for assessing pollutant effects on populations, i.e., traits that do not over- or underestimate population-level effects (Forbes et al. 2001; Forbes and

149

Calow 1999). In addition, whether a single trait is of low or high elasticity with respect to the population growth rate also depends on other factors, such as the life-cycle type of the species (e.g., semelparity or iteroparity; Calow et al. 1997). Based on an analysis of the published literature, Forbes et al. (2010) nonetheless found that in contrast to juvenile or adult survival and time to first reproduction fecundity tends to be a very sensitive life-cycle trait but it often exhibits low elasticity, i.e., it has a low impact on the population growth rate. To unravel the relationship between single life-cycle traits and population growth rate and to estimate pollutant effects on populations, full life-cycle experiments have been conducted in several studies (see references in Forbes et al. 2010; Forbes and Calow 1999; Sibly 1996). In most such studies, Daphnia spp. were used as the test organisms; however others focused on nematode life cycles (Alvarez et al. 2006, 2005; Kammenga et al. 1997; Kammenga and Riksen 1996; Kammenga et al. 1996; Nørhave et al. 2012; Vranken and Heip 1986; Wren et al. 2011). In the present study, a full life-cycle experiment (33 days) was carried out in order to examine the impact of an individual-level ivermectin lowest observed effect concentration (LOEC) on life-cycle traits of the nematode C. elegans. The LOEC (2.1 lg ivermectin/l) was determined for the endpoint reproduction in a previously conducted 4-day C. elegans bioassay (Brinke et al. 2011). As in the bioassay, component nematode growth gellan gum (CNGG) was used as the test medium as it has been shown to improve the accuracy of ivermectin testing (Brinke et al. 2011), which is a hydrophobic substance with a high sorption affinity (Prasse et al. 2009). Moreover, a hangingdrop method (adopted from Muschiol et al. 2009 with few modifications) was combined with the CNGG medium to conduct the life-cycle experiment. This facilitates ecotoxicological full life-cycle experiments with nematodes, inter alia, due to the high precision in determining life-cycle traits (discussed in Muschiol et al. 2009) and the flexible preparation of fresh test medium. The aim of the present study was to identify how the observed individual-level effects of ivermectin translate into population-level effects, and thus to determine whether the results from a bioassay can be applied in the protection of nematode populations exposed to ivermectin.

Methods Ivermectin A stock solution of ivermectin (Sigma, CAS 70288-86-7) dissolved in acetone was prepared and its concentration determined to be 424 lg/l. It was the identical stock solution used in the C. elegans bioassays of Brinke et al.

123

150

(2011) to achieve a final test concentration of 2.1 lg/l, the LOEC for the endpoint reproduction in the same study. The chemical analysis of ivermectin was based on the method reported by Krogh et al. (2008). Internal-standard solution was added to 400 ll of the stock solution and the volume made up to 1 ml with 10 mM ammonium formiate buffer (pH 4). This extract was analyzed by liquid chromatography-tandem mass spectrometry. Chromatographic separation was accomplished at ambient temperature using a Zorbax Eclipse XDB-C8 column (4.6 9 150 mm, 5 lm; Agilent Technologies, Waldbronn, Germany). A 50-ll sample of the extract was injected into the liquid chromatography system (Agilent 1100 with degasser, quaternary pump and autosampler, Agilent Technologies); the mobile phase was acetonitrile (A) and 10 mM ammonium formiate buffer (pH 4) (B). For all measurements, the tandem mass spectrometer (API 4000 with APCI ionization; Applied Biosystems, Foster City, USA) was operated in the positive ion mode using multiple reaction monitoring (ivermectin: m/z = 892.5/307 and 892.5/569.2). For further HPLC and mass spectrometry parameters, see Krogh et al. (2008). The limit of quantification was 5 lg/l. Test organism and test medium The hermaphrodite nematode C. elegans (strain N2) was provided by the Caenorhabditis Genetics Center (Minneapolis, USA) and maintained on nematode growth medium (NGM) agar (17 g agar, 2.5 g peptone, and 3 g NaCl in 975 ml deionized water, with 1 ml 1 M CaCl2, 1 ml 1 M MgSO4, 25 ml 1 M KPO4 buffer (pH 6), and 1 ml cholesterol solution (5 mg/ml in ethanol) added after autoclaving; Brenner 1974) seeded with Escherichia coli (OP50). The test medium was component nematode growth gellan gum (CNGG), a semi-fluid medium that uses gellan gum as the gelling agent instead of agar. CNNG was described by Brinke et al. (2011) as a test medium for nematode toxicity testing. Unlike agar, gellan-gum-based media solidify in the presence of (divalent) cations, such as magnesium or calcium, and can be re-liquefied by the addition of ethylenediaminetetraacetic acid (EDTA). The following medium components were separately prepared, autoclaved and stored at 4 °C until needed: (1) gellan gum solution (1.9 g/l GelriteÒ, Merck & Co., Inc., Kelco Division), with 1.25 ml/l cholesterol solution (5 mg/ml in ethanol) added after autoclaving; (2) salt solution (10 mM MgSO4, 10 mM CaCl2). Fresh medium was prepared every 2 days during the experimental period. Escherichia coli (OP50), as food bacteria, were suspended in sterile K-medium (3.1 g/l NaCl, 2.4 g/l KCl). The suspension was adjusted to a cell density of 6.25 9 109 cells/ml (final medium density:

123

M. Brinke et al.

approximately 5 9 109 cells/ml) and then centrifuged (20 min, 2,0009g). The supernatant was discarded and the pellet resuspended in gellan gum solution of the same volume. The final test medium was prepared in glass test tubes by adding 25 ll of the ivermectin stock solution directly to 4 ml gellan gum/bacterial suspension. Before mixing the medium with 0.5 ml salt solution, 475 ll of deionized water was added. Based on the dilution of the stock solution, C. elegans was exposed to an ivermectin concentration of 2.1 lg/l. For the control medium, 25 ll acetone was added instead of the stock solution. During the 2 days of use, the tubes were wrapped in aluminium foil, sealed with plastic paraffin film and stored at 4 °C. Experimental set-up For both the control and the ivermectin treatment, a cohort of 24 nematode individuals was followed throughout their entire lifespan. The experimental design was based on the study of Muschiol et al. (2009) with few modifications, such as the use of CNGG as medium and the avoidance of medium contact with plastic surfaces by using only glassware because of ivermectin’s high sorption affinity. Nematodes were kept individually in hanging drops of CNGG medium (12 ll) inside the lids of 12-well polystyrene multidishes. To avoid contact with the lids, each drop was placed on a round 18-mm glass cover slip previously mounted to the inside of the lid above each well. To prevent desiccation of the drops, each well contained a small piece of wet cellulose tissue and the multidishes were sealed with plastic paraffin film. The experiment was started with age-synchronized firststage (J1) juveniles that had hatched within 3.5 h. The nematodes were regularly transferred to fresh drops of CNGG medium (every 6–24 h) and survival was recorded. Nematodes were scored as dead if they showed a loss of turgor or did not respond to touch. Additionally, fecundity was recorded as the number of fertile eggs. Thus, the previous drop was incubated for 24 h until all fertile eggs had hatched. Then, the multidish lids were heated at 80 °C for one minute on a hot plate to relax the newborn juveniles and 12 ll of an EDTA/Rose Bengal solution (0.02 M EDTA, adjusted to pH 7 with NaOH, 300 mg/l Rose Bengal) was added to each warm drop to stain the juveniles and to liquefy the drop. Finally, a round 15-mm cover slip was placed on each drop and juveniles were counted under a dissecting microscope (25-fold magnification). Since fecundities were ascribed to the age-class in which the eggs were laid, the starting point of the experiment, and thus the age x = 0, was back-dated by the mean age of the introduced J1-juveniles (1.75 h) and the mean time eggs need for hatching after oviposition (7.3 ± 1.6 h; Muschiol et al. 2009). Thus, on average, the test organisms were defined to


be 9-h-old at the time they were transferred to the first drops. This measurement is important because age-specific fecundities have to be determined between two identical stages of successive generations (here egg to egg). The midpoint of each age-class was defined as the pivotal age. The length of each age-class was basically 24 h, but during the time of first egg oviposition and the main part of the reproductive period the accuracy for determining survival and fecundity was increased by using shorter lengths (age classes 2.46–3.22 days: 6 h; age classes 3.6–6.09 days: 12 h). Higher accuracy is needed particularly at the beginning of the reproductive period in order to more precisely determine life-cycle traits, such as the age at first reproduction. The experiment was conducted at 20 °C in the dark, except during the transfer of test organisms to fresh drops. The latter was done at room temperature behind a light shield in order to protect ivermectin from exposure to light. Individuals lost during transfer were excluded from the analysis (control: n = 2, ivermectin treatment: n = 0). Data analysis Based on the determined life tables and fecundity schedules, fitness, in terms of intrinsic rate of increase (rm), was estimated P rm x by iteration of the Euler–Lotka equation e lx mx ¼ 1, where lx is the age-specific survival probability to age-class x and mx is the age-specific fecundity (number of fertile eggs) of age-class x (Charlesworth 1994). This equation describes the growth rate of a population with a stable age-distribution that grows exponentially in an unlimited environment. The net P reproductive rate (R0), R0 ¼ lx mx (Charlesworth 1994), is the average number of offspring that an individual in a population will produce during its lifetime, depending on agespecific mortality rates. Based on lx, mx, rm, and R0, three alternative measures of generation time were calculated by the P P following equations: T0 ¼ xlx mx =R0 , T1 ¼ ðln R0 Þ=rm , P rm x and T ¼ xe lx mx , where T0 (also referred to as TC) is the mean age at reproduction of a cohort of females, T1 (also referred to as T) is the mean generation time, and T (also referred to as T) is the mean age of the mothers of a set of newborn individuals in a population with a stable age-distribution (Charlesworth 1994; Vranken and Heip 1983). For statistical analysis of rm, R0, T0, T1, and T a jack-knife procedure was used to derive pseudo values (Meyer et al. 1986). Differences in these pseudo values and in all other assessed life-cycle traits between the control and the ivermectin treatment were tested with Student’s t test (p \ 0.05) or, if the assumptions of normality and homogeneity of variance were not met, with the Mann–Whitney U test (p \ 0.05). The survival curves of the control and the ivermectin treatment were compared by using the log-rank-test (p \ 0.05). All statistical

151

operations were performed using the SPSS 15 for Windows statistical package (SPSS Inc.).

Results The ivermectin concentration of 2.1 lg/l inhibited the reproduction of C. elegans (Fig. 1). The total number of fertile eggs until the age-class 3.6 days (3.3–3.9 days), which approximately corresponds to the duration of the C. elegans bioassay (4 days; ISO 2010), was significantly reduced by 18.6 % (Table 1; Rbioassay). At the end of the experiment, the net reproductive rate (R0) was still an average of 12.4 % lower in the ivermectin treatment than in the control, but this difference was no longer significant (Table 1). The length of the reproductive period was one of the most sensitive life-cycle traits, as evidenced by the 25.9 % reduction determined in the ivermectin treatment (Table 1). Neither the beginning of the reproductive period (first oviposition) nor the age at maximum daily reproduction (Tmax rate) was significantly delayed due to ivermectin exposure (Table 1). Most differences in reproductive output between nematodes in the control and the ivermectin treatment occurred at the beginning of the reproductive period, with a significant reduction in the daily reproductive rates under the latter conditions (e.g., max rate; Table 1), whereas towards the end of the reproductive period the rates were similar to those of the control (Fig. 1). However, the daily reproductive rates of nematodes in age-classes 5.6–6.8 were slightly higher in the ivermectin treatment than in the control (Fig. 1). Furthermore, towards the end of the reproductive period, hatched juveniles in the ivermectin treatment were found

Fig. 1 Effects of ivermectin on the fecundity of C. elegans (mean ± SE); cohort sizes were n = 22 (control) and n = 24 (ivermectin treatment) at the start of the experiment. The n of the rate of fertile egg oviposition (per day) decreases in the control and the ivermectin treatment during the course of the experiment according to the corresponding survival (see Fig. 2)

123

152

M. Brinke et al.

Table 1 Summary of the ivermectin effects on life-cycle traits of C. elegans (mean ± SD), cohort sizes were n = 22 (control) and n = 24 (ivermectin treatment) at the start of the experiment Trait

Control

Ivermectin

% Inhibition

p value

Lifespan (d)

18.5 ± 8.03

12.95 ± 7.08

30.0

0.014

Juvenile period (d)

2.88 ± 0.15

2.89 ± 0.12

-0.5

n.s.

Repro period (d)

7.1 ± 1.98

5.26 ± 1.4

25.9

0.002

(d)

4.24 ± 0.57

4.32 ± 0.5

-2.0

n.s.

Max rate (N d-1)

121 ± 9

107 ± 12

11.2

0.001

R0 (N)

302 ± 32

265 ± 88

12.4

n.s.

93 ± 12

75 ± 27

18.6

0.007

3.5

0.011

Tmax

rate

Rbioassay (N) rm (d-1)

1.435 ± 0.037

1.385 ± 0.098

T0 (d)

4.59 ± 0.14

4.6 ± 0.19

-0.2

n.s.

T1 (d)

3.98 ± 0.11

4.03 ± 0.13

-1.3

n.s.

T (d)

3.61 ± 0.1

3.65 ± 0.12

-1.1

n.s.

Lifespan age at death, Juvenile period age at first fertile egg oviposition/minimum generation time, Repro period length of reproductive period, Tmax rate age at maximum daily rate of fertile egg oviposition, Max rate maximum daily rate of fertile egg oviposition, R0 net reproductive rate/total fertile eggs, Rbioassay total fertile eggs until age class corresponding to the C. elegans bioassay duration (4 days; ISO 10872), rm intrinsic rate of increase, T0, T1, T alternative measures of generation time, n.s. not significant

inside the body of nine individuals, while the mothers were still alive. This maternal hatching (endotokia matricida) resulted in the death of the mothers within 2 days after the observations were made. In contrast, in the control, hatched juveniles inside their mother’s body were found in two

Fig. 2 Effects of ivermectin on the survival of C. elegans; cohort sizes were n = 22 (control) and n = 24 (ivermectin treatment) at the start of the experiment. The black arrows indicate the time of the average start and end of the reproductive period in the control and the ivermectin treatment, while the white arrows indicate the end of the reproductive period of the longest reproducing individual

123

individuals, but the mothers were not alive at the time of their detection. The length of the post-reproductive period in the ivermectin treatment (5.7 ± 6.9 days) was shorter than in the control (8.5 ± 7.9 days) but this difference was not significant. There was an apparent effect of 2.1 lg ivermectin/l on the survival of C. elegans. This was apparent from the significant differences in the survival curves (Fig. 2). In addition, average lifespan was found to be the most sensitive life-cycle trait (30 % inhibition; Table 1), as nematodes exposed to ivermectin died approximately 5.5 days earlier than worms in the control. While there was no juvenile mortality in the control, in the ivermectin treatment two individuals had not exceeded the second juvenile stage at the time of their death after 2.5 and 4.1 days (8.3 % of the observed cohort). The maximum age reached by a single individual in the control and in the ivermectin treatment was 32.8 and 27.8 days, respectively. Moreover, ivermectin-treated individuals, who died after the end of reproductive period, were much less active, including minimal movement, than the controls for many more days preceding their death. Although significantly different from the control, the intrinsic rate of increase was only 3.5 % lower in the ivermectin treatment (Table 1). All calculated generation times were only marginally and not significantly prolonged due to ivermectin exposure (Table 1).

Discussion The full life-cycle experiment carried out in this work reproduced the effects of ivermectin on fecundity determined in a previous 4-day bioassay (Brinke et al. 2011). The measurement used in this study, the total number of fertile eggs until age-class 3.6 days (3.3–3.9 days; Rbioassay), is comparable to the bioassay’s endpoint reproduction (offspring per test organism). Although in the life-cycle experiment time was measured from a freshly laid egg and not, as in the bioassay, from a freshly hatched J1-juvenile, the two measurements correspond because in the bioassay only hatched juveniles were counted and the present fertile eggs were ignored; that is, the counted juveniles originated from fertile eggs that were laid at about 7.3 h (mean egg hatching time; see Methods) before the bioassay ended. Moreover, in both experiments the nematodes were exposed to ivermectin beginning at the J1-juvenile stage. At a concentration of 2.1 lg ivermectin/l there was a significant reduction of the Rbioassay of 18.6 %, while in the bioassay of Brinke et al. (2011) reproduction was significantly reduced by 22.0 %. Thus, the two experimental approaches detected a similar impact of ivermectin on the early fecundity of C. elegans. However, the impact on total fecundity, in terms of the net


reproductive rate (R0), did not intensify until the end of the C. elegans life cycle. Instead, it was comparatively weaker and, in contrast to the Rbioassay, the R0 of the ivermectin treatment was not significantly different from the control value. Nonetheless, it was still reduced, by an average of 12.4 %. Moreover, the higher variability in the reproductive output of the ivermectin-exposed nematodes than in the control nematodes could have masked significant effects (see Fig. 1). In conclusion, while the bioassay approach seems to be sufficiently protective with respect to the effects of ivermectin on C. elegans fecundity, the life-cycle experiment revealed other, more sensitive traits, such as average lifespan and length of the reproductive period (mean inhibition of 30 and 25.9 %, respectively). The increased maternal hatching in the ivermectin treatment contributed to the reduced reproductive period and, as a consequence, to the earlier deaths of the nematodes. It is known that pollutants can induce the phenomenon of maternal hatching (endotokia matricida). For example, Nørhave et al. (2012) observed that the frequency of maternal hatching increased with increasing cadmium concentrations. In the present study, the maternal hatching may be particularly attributed to ivermectin’s neurotoxic mode of action, likely resulting in impaired gonadal contraction waves and hence oocyte transport in the worm (Bernt et al. 1998). Moreover, this specific mode of action also explains the decreased activity and mobility prior to the earlier deaths of the longest-living individuals in the ivermectin treatment, because ivermectin paralyzes the nematodes’ pharyngeal and somatic muscles, an effect that is ultimately fatal (Duce and Scott 1985; Geary et al. 1993). This reduced activity of the nematodes per se may have also affected the oocyte transport. These effects may influence the viability of populations in the environment because they indicate a reduced potential to delay reproduction to compensate for the lower (daily) reproduction rates. Moreover, the considerable effects of a low ivermectin concentration on survival under laboratory conditions raise concerns about the vulnerability of ivermectinexposed populations to additional environmental impacts, such as predation. Jager and Klok (2010) stressed that background mortality should be considered when extrapolating the effects of life-cycle experiments to populations in the environment, since under field conditions mortality rates are likely much higher. In the present study, a significant negative effect of the bioassay-derived ivermectin LOEC on the population growth rate could be demonstrated, but the populationlevel response was less sensitive to ivermectin than the responses of significantly affected life-cycle traits (rm \ max rate \ Rbioassay \ repro period \ lifespan; see Table 1). In studies investigating the effects of cadmium and pentachlorophenol on the life-cycle traits of the nematode Plectus acuminatus, the duration of the reproductive period was, as

153

in the present study, one of the most sensitive traits, but it had no effect on the population growth rate (Kammenga et al. 1997, 1996). Forbes et al. (2010) demonstrated that in many studies survival, and particularly juvenile survival, had a higher elasticity than fecundity, although in contrast to the present study survival often was a less sensitive trait than fecundity. In conclusion, it is likely that in this work the considerable effects of ivermectin on the high-elastic trait survival, including a juvenile mortality of 8.3 % of the cohort, contributed more to the observed population-level response than did reductions in the reproductive period or fecundity. Thus, if the negative correlation between sensitivity and trait elasticity, as shown by Forbes et al. (2010), is not consistently valid for nematodes exposed to ivermectin, as demonstrated herein, then the pharmaceutical will place the non-target nematode populations at higher risk. However, although our results show that a high-elastic trait was sensitive towards ivermectin, the percentage decrease in the population growth rate was not very high. A similar effect was reported by Kammenga et al. (1997) in a study in which exposure to pentachlorophenol reduced the high-elastic trait juvenile survival of the nematode P. acuminatus by 20 %, but this led to only a 5 % decrease of the population growth rate. However, although in that study survival was more strongly affected, it still was less sensitive in its response than other life-cycle traits, but the trait was identified to be the most important contributor to the reduction of the growth rate. While population growth rate has been criticized as being a less sensitive response trait to pollutants than other traits, it may be a more ecologically relevant and realistic endpoint (Walthall and Stark 1997). Moreover, it should be considered that, in the long term, small changes in the intrinsic rate of increase per day might lead to larger population effects (Kammenga et al. 1997) and that effects might get more pronounced in the following generations (Derycke et al. 2007; Lira et al. 2011).

Conclusion A 4-day bioassay may be protective, if not overprotective, regarding ivermectin’s effects on fecundity of C. elegans, because, as shown in the present study, the impact on fecundity is diminished rather than intensified during the remaining reproductive period. Furthermore, while our lifecycle experiment was able to prove a significant effect of the bioassay-derived LOEC on the population level, the results indicated that fecundity may not be the crucial contributor to the population-level response, i.e., the decreased population growth rate. Although fecundity and length of the reproductive period were also reduced in the life-cycle experiment, effects on survival may have contributed more to the

123

154

population-level response. Thus, in the present life-cycle experiment the extrapolation uncertainty of ivermectin’s effects on the population level could be reduced because of the integrated consideration of several life-cycle traits, whereas the bioassay showed a sensitive response of a lifecycle trait that might not immediately affect population viability. In conclusion, the present study generally stresses the importance of studies that extrapolate effects to the population level with lower uncertainties. Furthermore, the life-cycle experiment complements studies of ivermectin effects on free-living nematodes at the organismic and community level and underlines that populations of freeliving nematodes, which are closely related to the main target organisms of the drug and which are of great ecological importance in sediments and soils, are likely vulnerable to ivermectin exposure in the environment. Acknowledgments This study was funded by the Federal Ministry of Transport, Building and Urban Development, Berlin and Bonn, Germany. We thank Guido Fink (Federal Institute of Hydrology, Koblenz, Germany) for analyzing the ivermectin stock solution. The C. elegans strain (N2) was provided by the Caenorhabditis Genetics Center (Minneapolis, MN, USA), which is funded by the National Institutes of Health—National Center for Research Resources (USA).

References Alvarez OA, Jager T, Kooijman SALM, Kammenga JE (2005) Responses to stress of Caenorhabditis elegans populations with different reproductive strategies. Funct Ecol 19:656–664 Alvarez OA, Jager T, Redondo EM, Kammenga JE (2006) Physiological modes of action of toxic chemicals in the nematode Acrobeloides nanus. Environ Toxicol Chem 25:3230–3237 Ardelli BF, Stitt LE, Tompkins JB, Prichard RK (2009) A comparison of the effects of ivermectin and moxidectin on the nematode Caenorhabditis elegans. Vet Parasitol 165:96–108 Arena JP, Liu KK, Paress PS, Frazier EG, Cully DF, Mrozik H, Schaeffer JM (1995) The mechanism of action of avermectins in Caenorhabditis elegans: correlation between activation of glutamate-sensitive chloride current, membrane binding, and biological activity. J Parasitol 81:286–294 Bernt U, Junkersdorf B, Londershausen M, Harder A, Schierenberg E (1998) Effects of anthelminthics with different modes of action on the behavior and development of Caenorhabditis elegans. Fundam Appl Nematol 21:251–263 Bongers T, Ferris H (1999) Nematode community structure as a bioindicator in environmental monitoring. Trends Ecol Evol 14:224–228 Brenner S (1974) Genetics of Caenorhabditis elegans. Genetics 77:71–94 Brinke M, Ho¨ss S, Fink G, Ternes TA, Heininger P, Traunspurger W (2010) Assessing effects of the pharmaceutical ivermectin on meiobenthic communities using freshwater microcosms. Aquat Toxicol 99:126–137 Brinke M, Heininger P, Traunspurger W (2011) A semi-fluid gellan gum medium improves nematode toxicity testing. Ecotoxicol Environ Saf 74:1824–1831 Calow P, Sibly RM, Forbes V (1997) Risk assessment on the basis of simplified life-history scenarios. Environ Toxicol Chem 16: 1983–1989

123

M. Brinke et al. Campbell WC, Fisher MH, Stapley EO, Albers-Scho¨nberg G, Jacob TA (1983) Ivermectin: a potent new antiparasitic agent. Science 221:823–828 Charlesworth B (1994) Evolution in age-structured populations, 2nd edn. Cambridge University Press, Cambridge Derycke S, Hendrickx F, Backeljau T, D’Hondt S, Camphijn L, Vincx M, Moens T (2007) Effects of sublethal abiotic stressors on population growth and genetic diversity of Pellioditis marina (Nematoda) from the Westerschelde estuary. Aquat Toxicol 82:110–119 Duce IR, Scott RH (1985) Actions of dihydroavermectin B1a on insect muscle. Br J Pharmacol 85:395–401 Edwards CA, Atiyeh RM, Ro¨mbke J (2001) Environmental impact of avermectins. Rev Environ Contam Toxicol 171:111–137 Ferson S, Ginzburg LR, Goldstein RA (1996) Inferring ecological risk from toxicity bioassays. Water Air Soil Pollut 90:71–82 Forbes VE, Calow P (1999) Is the per capita rate of increase a good measure of population-level effects in ecotoxicology? Environ Toxicol Chem 18:1544–1556 Forbes VE, Calow P, Sibly RM (2001) Are current species extrapolation models a good basis for ecological risk assessment? Environ Toxicol Chem 20:442–447 Forbes VE, Calow P, Sibly RM (2008) The extrapolation problem and how population modeling can help. Environ Toxicol Chem 27:1987–1994 Forbes VE, Olsen M, Palmqvist A, Calow P (2010) Environmentally sensitive life-cycle traits have low elasticity: implications for theory and practice. Ecol Appl 20:1449–1455 Geary TG, Sims SM, Thomas EM, Vanover L, Davis JP, Winterrowd CA, Klein RD, Ho NFH, Thompson DP (1993) Haemonchus contortus: ivermectin-induced paralysis of the pharynx. Exp Parasitol 77:88–96 ISO (2010) Water quality—Determination of the toxic effect of sediment and soil samples on growth, fertility and reproduction of Caenorhabditis elegans (Nematoda). International Organization for Standardization 10872. Geneva, Switzerland Jager T, Klok C (2010) Extrapolating toxic effects on individuals to the population level: the role of dynamic energy budgets. Philos Trans R Soc Lond B Biol Sci 365:3531–3540 Kammenga JE, Riksen JAG (1996) Comparing differences in species sensitivity to toxicants: phenotypic plasticity versus concentration-response relationships. Environ Toxicol Chem 15:1649–1653 Kammenga JE, Busschers M, Van Straalen NM, Jepson PC, Bakker J (1996) Stress induced fitness reduction is not determined by the most sensitive life-cycle trait. Funct Ecol 10:106–111 Kammenga JE, van Koert PHG, Koeman JH, Bakker J (1997) Fitness consequences of toxic stress evaluated within the context of phenotypic plasticity. Ecol Appl 7:726–734 Krogh KA, Bjo¨rklund E, Lo¨ffler D, Fink G, Halling-Sørensen B, Ternes TA (2008) Development of an analytical method to determine avermectins in water, sediments and soils using liquid chromatography-tandem mass spectrometry. J Chromatogr A 1211:60–69 Liebig M, Alonso Fernandez A, Blu¨baum-Gronau E, Boxall A, Brinke M, Carbonell G, Egeler P, Fenner K, Fernandez C, Fink G, Garric J, Halling-Sørensen B, Knacker T, Krogh KA, Ku¨ster A, Lo¨ffler D, Porcel Cots MA, Pope L, Prasse C, Ro¨mbke J, Ro¨nnefahrt I, Schneider MK, Schweitzer N, Tarazona JV, Ternes TA, Traunspurger W, Wehrhan A, Duis K (2010) Environmental risk assessment of ivermectin: a case study. Integr Environ Assess Manage 6:567–587 Lira VF, Santos GAP, Derycke S, Larrazabal MEL, FonseˆcaGenevois VG, Moens T (2011) Effects of barium and cadmium on the population development of the marine nematode Rhabditis (Pellioditis) marina. Mar Environ Res 72:151–159 Meyer JS, Ingersoll CG, McDonald LL, Boyce MS (1986) Estimating uncertainty in population growth rates: jackknife vs. bootstrap techniques. Ecology 67:1156–1166

Effects of a bioassay-derived ivermectin on C. elegans Muschiol D, Schroeder S, Traunspurger W (2009) Life cycle and population growth rate of Caenorhabditis elegans studied by a new method. BMC Ecol 9:14 Nørhave NJ, Spurgeon D, Svendsen C, Cedergreen N (2012) How does growth temperature affect cadmium toxicity measured on different life history traits in the soil nematode Caenorhabditis elegans? Environ Toxicol Chem 31:787–793 Omura S (2008) Ivermectin: 25 years and still going strong. Int J Antimicrob Agents 31:91–98 Prasse C, Lo¨ffler D, Ternes TA (2009) Environmental fate of the anthelmintic ivermectin in an aerobic sediment/water system. Chemosphere 77:1321–1325 Sibly RM (1996) Effects of pollutants on individual life histories and population growth rates. In: Newman MC, Jagoe CH (eds) Ecotoxicology: a hierarchical treatment. CRC/Lewis Publishers, Boca Raton, pp 197–223 Traunspurger W, Michiels IC, Abebe E (2006) Composition and distribution of free-living freshwater nematodes: global and local perspectives. In: Abebe E, Andrassy I, Traunspurger W (eds) Freshwater nematodes: ecology and taxonomy. CABI Publishing, Wallingford, pp 46–76 Van Leeuwen CJ, Luttmer WJ, Griffioen PS (1985) The use of cohorts and populations in chronic toxicity studies with Daphnia magna: a cadmium example. Ecotoxicol Environ Saf 9:26–39 Vranken G, Heip C (1983) Calculation of the intrinsic rate of natural increase, rm, with Rhabditis marina Bastian 1865 (Nematoda). Nematologica 29:468–477

155 Vranken G, Heip C (1986) Toxicity of copper, mercury and lead to a marine nematode. Mar Pollut Bull 17:453–457 Walthall WK, Stark JD (1997) Comparison of two population-level ecotoxicological endpoints: the intrinsic (rm) and instantaneous (ri) rates of increase. Environ Toxicol Chem 16:1068–1073 Wren JF, Kille P, Spurgeon DJ, Swain S, Sturzenbaum SR, Jager T (2011) Application of physiologically based modelling and transcriptomics to probe the systems toxicology of aldicarb for Caenorhabditis elegans (Maupas 1900). Ecotoxicology 20:397– 408 Yeates GW, Bongers T, de Goede RGM, Freckman DW, Georgieva SS (1993) Feeding habits in soil nematode families and genera— an outline for soil ecologists. J Nematol 25:315–331 Yeates GW, Dimander SO, Waller P, Hoglund J (2003) Soil nematode populations beneath faecal pats from grazing cattle treated with the ivermectin sustained-release bolus or fed the nematophagous fungus Duddingtonia flagrans to control nematode parasites. Acta Agric Scand A: Anim Sci 53:197–206 Yeates GW, Hrabok JT, Oksanen A, Nieminen M, Waller PJ (2007a) Soil nematode populations beneath faeces from reindeer treated with ivermectin. Acta Agric Scand B: Soil Plant 57:126–133 Yeates GW, Skipp RA, Gray RAJ, Chen LY, Waghorn TS (2007b) Impact on soil fauna of sheep faeces containing a range of parasite control agents. Appl Soil Ecol 35:380–389

123