Toxicity of imidacloprid to the terrestrial isopod Porcellio scaber (Isopoda, Crustacea)

Available online at www.sciencedirect.com

Chemosphere 71 (2008) 1326–1334 www.elsevier.com/locate/chemosphere

Toxicity of imidacloprid to the terrestrial isopod Porcellio scaber (Isopoda, Crustacea) Damjana Drobne a,*, Mateja Blazˇicˇ b, Cornelis A.M. Van Gestel c, Vladka Lesˇer a, Primozˇ Zidar a, Anita Jemec d, Polonca Trebsˇe b a

University of Ljubljana, Biotechnical Faculty, Department of Biology, Vecˇna Pot 111, SI-1000 Ljubljana, Slovenia University of Nova Gorica, Laboratory for Environmental Research, P.O. Box 301, SI-5001 Nova Gorica, Slovenia c Vrije Universiteit, Department of Animal Ecology, Institute of Ecological Science, De Boelelaan 1085, NL-1081 HV Amsterdam, The Netherlands d National Institute of Chemistry, Hajdrihova 19, SI-1000 Ljubljana, Slovenia b

Received 6 July 2007; received in revised form 5 November 2007; accepted 15 November 2007 Available online 10 January 2008

Abstract Imidacloprid is a neonicotinoid insecticide with neurotoxic action that, as a possible alternative for commonly used organophosphorus pesticides, has gained registration in about 120 countries for use in over 140 agricultural crops. Only few data are available on its toxicity for soil invertebrates. We therefore assessed the effects of imidacloprid on survival, weight gain, feeding rate, total protein content, glutathione S-transferase activity (GST), and digestive gland epithelial thickness in juveniles and adults of the terrestrial isopod Porcellio scaber. After two weeks of feeding on imidacloprid-dosed food, weight gain (NOEC 5 lg/g dry food) and feeding rate (NOEC 10 lg/g) in juveniles, and feeding rate (NOEC < 10 lg/g) and digestive gland epithelial thickness (NOEC < 10 lg/g) in adults were most affected. In juveniles induction of GST activity and increase of total protein content per wet animal weight was detected at 5 lg/g dry food, whereas in adults a reduction of GST was observed at 25 lg/g (NOEC 10 lg/g). An estimate of actual intake rates suggests that imidacloprid affects isopods at similar exposure concentrations as insects. The toxicity of imidacloprid was similar to that of the organophosphorus pesticide diazinon, tested earlier using the same methods [Stanek, K., Drobne, D., Trebse, P., 2006. Linkage of biomarkers along levels of biological complexity in juvenile and adult diazinon fed terrestrial isopod (Porcellio scaber, Isopoda, Crustacea). Chemosphere 64, 1745–1752]. At actual environmental concentrations, diazinon poses a higher risk to P. scaber. Due to its increasing use in crop protection and higher persistence in soil, imidacloprid might however, be potentially more dangerous after long-term application. We conclude that toxicity testing with P. scaber provides relevant, repeatable, reproducible and comparable toxicity data that is useful for the risk assessment of pesticides in the terrestrial environment. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Biomarkers; Toxicity testing; Glutathione S-transferase; Digestive gland epithelium; Risk quotient; Environmental risk characterization

1. Introduction The market for pesticides is changing as a result of the US EPA phase out of most urban uses of two of the most commonly used organophosphorus (OP) pesticides, diazinon and chlorpyrifos (TDC Environmental, 2003). However, agricultural uses of diazinon continued after December 2004 in a wide range, what led the US EPA to *

Corresponding author. Tel.: +386 1 42 33 388; fax: +386 1 25 73 390. E-mail address: [email protected] (D. Drobne).

0045-6535/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2007.11.042

conduct a cumulative risk assessment of all OPs. The US EPA has reached the decision that diazinon poses unacceptable risk to agricultural workers, birds and other wildlife species (Cobb et al., 2000). The situation is similar in the European Union and will probably lead to a progressive withdrawal of diazinon from use also in Europe. In the UK, diazinon is being progressively phased out because of a lack of data to support its continued use (APVMA, 2003). The US EPA created a candidate list of ten alternatives to diazinon (TDC Environmental, 2003). Among them,

D. Drobne et al. / Chemosphere 71 (2008) 1326–1334

imidacloprid succeeded to have the world’s fastest growing sales (TDC Environmental, 2003). Imidacloprid belongs to a major new class of insecticides, called neonicotinoids, which are accounting for 11– 15% of the total insecticide market (TDC Environmental, 2003). Since its launch in 1991, products containing imidacloprid have gained registrations in about 120 countries and are marketed for use in agriculture (for over 140 agricultural crops), on turf, on pets, and for household pests (Liu et al., 2005). Imidacloprid is marketed under variety of names including Gaucho, Merit, Admire, Confidor and Winner. Although imidacloprid has been in use for a relatively short period of time compared to other common pesticides, it is considered to being used in the largest volume globally of all insecticides (Cox, 2001; Ware and Whitacre, 2004). Imidacloprid can be used for seed dressing, foliar application, trunk injection, soil injection and soil drench. Application rates of imidacloprid for different crops range from 0.05 to 0.4 kg of active ingredient/ha (Bayer Technical Information and ConfidorÒ, 2000), while 0.03 kg a.i./ha is the recommended application rate for tea fields in India (Anatra-Cordone and Durkin, 2005). Imidacloprid is a nicotinic acetylcholine receptor (nAChR) agonist (Buckingham et al., 1997). Its mechanism of action has been studied extensively, and is relatively well known. In essence, imidacloprid activates nAChR through binding at or near the sites where nicotine and acetylcholine bind, resulting in dysfunction of the nervous system, immobilization or death (Anatra-Cordone and Durkin, 2005). Imidacloprid has different affinities to nAChRs of different organisms. The selectivity of vertebrate and invertebrate nAChR is attributed to structural heterogeneity of neuronal nAChRs and leads to differences in the sensitivity to imidacloprid (Lindstrom et al., 1995; Lind et al., 2001; Matsuda et al., 2001; Tomizawa and Casida, 2005). Available data indicate that imidacloprid can be highly toxic to some aquatic crustaceans, but generally less toxic to fish (TDC Environmental, 2003; Jemec et al., 2007a). Also, the LD50s for mammals and birds are much higher than for invertebrates (Anatra-Cordone and Durkin, 2005). Due to the high insecticidal activity and low mammalian toxicity, imidacloprid is considered safer compared to traditional OP, carbamate and pyrethroid insecticides (Felsot and Ruppert, 2002). In addition imidacloprid can be applied at very low rates. However, very little data are available on the toxicity of imidacloprid to non-target organisms, especially those inhabiting terrestrial ecosystems. A few comparative toxicity studies indicate high species-specific response to imidacloprid, which suggests that imidacloprid toxicity data may not be generalized. Terrestrial isopods may be suitable test organisms due to their well-known biology and physiology, relative ease of laboratory maintenance, and possibility to acquire individual toxicity data. Also, due to their important ecological role as decomposers of organic material, terrestrial isopods are widely accepted as test organisms in terrestrial ecotoxi-

1327

cology and ecophysiology (Drobne, 1997; Løkke and Van Gestel, 1998; Lapanje et al., 2007). Isopods also allow measuring biomarkers at different levels of biological complexity. Biomarkers may be divided into biomarkers of exposure and biomarkers of effect. By a definition posed by Kammenga et al. (2000), biomarkers of effect are those measured alterations in an organism that are associated with possible health impairment or disease. Biomarkers of exposure are those measured responses of an organism that indicate the interaction between a xenobiotic agent and some target molecule or cell (Roberts and Oris, 2004). Biomarkers of exposure in our study are S-transferase (GST) activity and total protein content per wet animal weight and biomarkers of effects are hepatopancreas epithelial thickness, feeding rate, weight change and mortality. The aim of our work was: (a) To assess toxicity of imidacloprid in two-week laboratory toxicity tests by analyzing biomarkers at different levels of biological complexity in juvenile and adult terrestrial isopod Porcellio scaber; (b) To compare toxicity data on imidacloprid and toxicity data on diazinon tested in the same experimental set-up; (c) To discuss relative toxicity of imidacloprid to diazinon on the basis of data obtained in our study and those reported in the literature; (d) To evaluate the potential of the toxicity testing protocol with the terrestrial isopod P. scaber test system for providing toxicity data for risk characterization of pesticides in the terrestrial environment.

2. Materials and methods 2.1. Chemicals Imidacloprid (99.8%) was provided by Bayer Crop Science Slovenia. Sodium hydrogen phosphate (Na2HPO4) and potassium dihydrogen phosphate (KH2PO4) were purchased from Fluka. Bovine serum albumin (albumin fraction V) was purchased from Merck, Bradford reagent, 1–chloro-2,4-dinitrobenzene (CDNB), and L-glutathione (reduced form) (GSH) were purchased from Sigma Aldrich and concentrated sulphuric acid from Carlo Erba. Methanol (HPLC grade), abs. ethanol and chloroform were obtained from J.T. Baker. 2.2. Test organisms Terrestrial isopods (P. scaber, Isopoda, Crustacea) were collected from the litter layer of uncontaminated woodlands near Nova Gorica (Slovenia) and Ljubljana. In the laboratory, the animals were kept in a terrarium (20  35  20 cm) filled with a layer of moistened sand and soil (2–5 cm) and a thick layer of partly decomposed hazelnut tree leaves (Corylus avellana). The substratum in

1328

D. Drobne et al. / Chemosphere 71 (2008) 1326–1334

the terrarium was heated up to 80 °C for several hours to destroy predators (spiders) before introduction of the isopods. Hazelnut tree leaves were collected from uncontaminated woodland and air-dried at room temperature. The culture was sprayed daily with water and kept at room temperature and high humidity up to three weeks before the experiments. 2.3. Experimental set-up Three two-week experiments were conducted, in which juvenile or adult isopods were fed imidacloprid-treated food. In all experiments we determined mortality, faecal pellets production (defecation) as a measure of feeding activity and weight change of exposed juveniles and adults. In addition to these measures, in one of the experiments on adults and the experiment on juveniles, total protein content and GST activity were analyzed, while in the second experiment on adults epithelial thickness of the digestive gland (hepatopancreas) was determined. Adults with body weights ranging from 30 to 64 mg, of both sexes, were fed with hazelnut tree leaves (C. avellana) dosed with imidacloprid. Leaves were dosed with pesticide as described by Stanek et al. (2006). Dry leaves were cut into pieces of approximately the same surface area and a weight of 100 mg. From solutions of different concentrations of imidacloprid, prepared from a stock solution in distilled water, 150 ll was applied to the lower leaf surface as small droplets and evenly spread over the surface with a paintbrush. The nominal concentrations of imidacloprid in the leaves were 10 and 25 lg imidacloprid/g dry food. Each animal was placed individually in a Petri dish, to which pieces of leaves dosed with imidacloprid were added. Humidity in the Petri dishes was maintained by regular spraying distilled water on the internal side of the lids. Juveniles with body weights from 12.5 to 30 mg, of both sexes, were fed with briquettes prepared as proposed by Kaschl et al. (2002) and Zidar et al. (2005). Dry leaves were ground with a coffee mill and 0.25 mm mesh sieved. Leaf powder, gelatine, and a fish food were mixed in a 63:34:3 ratio and turned into a paste by adding demineralised water (15 ml per g of gelatine). For the preparation of imidacloprid-dosed food, imidacloprid was added to the paste to give concentrations 0, 2.5, 5, 10, and 50 lg/g dry food. Each animal was placed individually in a Petri dish, with two pieces of briquettes, dosed with imidacloprid. All Petri dishes were placed in a large plastic covered container maintained at relative humidity close to 100%. Every third day the remaining food was removed, replaced by fresh leaf pieces or briquettes, air-dried and weighed. The animals were weighed at the start and end of the experiment. Every seven days (±1 day) faecal pellets were removed, dried and weighed. In the experiment with juveniles, 10 animals per treatment were exposed. At the end of the exposure period we determined mortality, weight gain, feeding rate, total protein content and GST activity.

With adults three different experiments were conducted. Here, altogether 36 control animals were exposed, 42 animals fed with 10 lg/g of imidacloprid, and 22 animals fed with 25 lg/g of imidacloprid. Data obtained in the three different experiments are combined and presented together, since the controls were not statistically significantly different. In all groups we analyzed mortality, weight change and feeding rate. In addition, part of the animals were analysed for biochemical or histological parameters. 2.4. Determination of glutathione S-transferase activity The measurement of the GST activity was performed with 1-chloro-2,4-dinitrobenzene as a substrate according to Habig et al. (1974) and Jemec et al. (2007b). A single animal was homogenized in potassium phosphate buffer (100 mM, pH 6.5) (1.2 ml) and centrifuged at 3000 rpm for 15 min at 0 °C. The blank solution of a reaction mixture containing 900 ll of phosphate buffer (100 mM, pH 6.5), 25 ll of 1-chloro-2,4-dinitrobenzene solution (60 mM) and 50 ll of reduced glutathione (40 mM) was followed spectrophotometrically (Spectrophotometer HP 8453, USA) at 340 nm and room temperature for 3 min. Afterwards 25 ll of supernatant was added and the reaction was followed for another 5 min. The blank reaction was subtracted from the measurement containing supernatant. Samples were measured in duplicate. Enzyme activities were expressed in lmoles of conjugated GSH/min, mg of total protein (extinction coefficient e340 = 9600 M 1 cm 1). 2.5. Analysis of total proteins Total proteins were determined in all surviving animals. For protein determination, 50 ll of the supernatant of a single animal’s tissue was added to 1500 ll Bradford reagent (Bradford, 1976). Two replicates were made for each animal sample. The protein standards were made with bovine serum albumin (BSA). The absorbance of the blue colored complex was measured spectrophotometrically at 595 nm. The total protein content was calculated as mg of protein/mg animal wet weight. 2.6. Epithelial thickness All four digestive gland tubes of the hepatopancreas were fixed in Carnoy B fixative (ethanol:chloroform:acetic acid; 6:3:1) for 2 h at room temperature. The fixed digestive gland specimens were dehydrated in ethanol, cleared in xylene and embedded in Paraplast Plus (Sigma). Eightmicrometer sections (Reichert-Jung 2040 rotatory microtome, Reichert-Jung, Vienna, Austria) of the entire digestive gland tube were cut and stained with eosin (Drobne and Drobne, 2005). The sections were examined with a light microscope (Carl Zeiss Axioskop 2 MOT, Carl Zeiss, Jena, Germany). For the determination of the average epithelial thickness 12 sections equally apart along one gland

D. Drobne et al. / Chemosphere 71 (2008) 1326–1334

tube were chosen and analyzed. The average epithelial thickness on a section was determined using the image analysis computer program KS 400 (Carl Zeiss Vision, Carl Zeiss Jena GmbH, Oberkochen, Germany) and manual contouring of the inner and outer epithelial surface (Drobne and Drobne, 2005). 2.7. Analysis of imidacloprid concentrations in food Imidacloprid concentrations in leaves and briquettes at the start and end of the two-week exposure periods were analyzed using a high performance liquid chromatograph (HPLC, Agilent 1100) with a diode array detector. HPLC conditions were as follows: Zorbax C8 (25 cm  4.6 mm, 53 lm) column coupled with Alltech precolumn. The mobile phase consisted of deionised water and acetonitrile (80:20). In case of the briquettes, the mobile phase was amended with 2.5% phosphoric acid to avoid interference with the matrix. Separation was performed in isocratic mode with a flow rate of 1.25 ml/min. The temperature was maintained at 25 °C during the analysis. The retention time for imidacloprid on leaves was 12 min, while it was 15 min for briquette samples. The difference in retention times is due to the different solvent systems applied to avoid matrix–signal overlap. Leaf samples were prepared as follows: 500 mg of contaminated leaves (10 and 25 mg/kg) were homogenized and extracted with 20 ml of acetonitrile. After 10 min of extraction, the acetonitrile was decanted. The extraction was repeated twice with 15 ml of acetonitrile. The organic phases were combined and evaporated to dryness. An aliquot of water (5 ml) was then added, the mixture was solid phase extracted using C18 cartridges, activated with 1 ml of acetonitrile. Briquette samples were prepared as follows: 1 g of briquettes, containing different amounts of imidacloprid were homogenized, extracted three times with 10 ml ethyl acetate and filtered through cellulose filters. After evaporation of the solvent, the residue was dissolved in 1 ml of mobile phase and prepared for HPLC analysis. The R2 value for linear regression of the calibration curve was 0.9904 and 0.9994 for leaves and briquette analyses, respectively. The relative standard deviation in all experiments was 5–15%. The extraction efficiency was determined on the basis of the calibration curve for water samples. For leaf concentrations ranging from 0.5 to 10 lg/g recovery was 93%, while it was 18% for briquette concentrations ranging from 2.5 to 50 lg/g. The low extraction efficiency for briquettes may be explained from the high sugar content, which makes this a much more complex matrix. 2.8. Statistical analyses of data The difference of the medians of measured parameters in exposed and unexposed groups was tested applying the non-parametric Mann–Whitney test. The lowest concentration of pesticide that caused significantly different

1329

responses of animals from that in the control group was considered as the lowest observed effect concentration (LOEC) of pesticide, the highest tested concentration at which no significant differences from control were found was the no-observed effect concentration (NOEC). All calculations were done using SPSS 14.0.1 for Windows statistic software. Using the amount of food consumed, imidacloprid intake rates could be calculated. These data were used to convert NOEC and LOEC values into no-observed or lowest-observed effect levels (NOEL and LOEL, respectively). 3. Results We determined the concentrations of imidacloprid on leaves and briquettes at the beginning as well as the end of the experiment. Measured concentrations were within 10% of the nominal ones. There was no decrease in imidacloprid content (concentrations 10 and 25 lg/g dry wt. imidacloprid of food) applied on leaves and left for up to two weeks to the conditions present during the experiment with P. scaber. All effects are expressed on the basis of nominal concentrations. Tested concentrations of imidacloprid did not cause mortality of juveniles or adults. However, when animals were disturbed while they were moulting, they died irrespective of the treatment. In case of juveniles, one out of 10 died in a group exposed to 10 lg/g (10%). In the tests with adults, mortality amounted five out of 36 (14%), three out of 42 (7.1%) and one out of 22 (4.5%), respectively in the controls, and the 10 lg/g and 25 lg/g imidacloprid exposure groups. Data on the toxicity of imidacloprid to juvenile and adult P. scaber after two-weeks of feeding exposure are presented in Table 1. Average (±SD) feeding rate of juveniles in a control group was 0.059 ± 0.01 g food/g wet wt./day, in a group exposed to 2.5 lg/g imidacloprid average feeding rate was 0.067 ± 0.013; in a group exposed to 5 lg/g it was 0.059 ± 0.009 g food/g wet wt./day; in a group exposed to 10 lg/g it was 0.052 ± 0.012 g food/g wet wt./day and in a group fed with 50 lg/g, feeding rate was 0.040 ± 0.009 g food/g wet wt./day. In juveniles, feeding rate was statistically significantly reduced at 50 lg/g imidacloprid (Table 1). Average feeding rate of adults in a control group was 0.043 ± 0.012 g food/g wet wt./day, in a group exposed to 10 lg/g it was 0.023 ± 0.012 g food/g wet wt./day and in a group exposed to 25 lg/g, feeding rate was 0.014 ± 0.005 g food/g wet wt./day. In adults, feeding rate was statistically significantly reduced at 10 lg/g imidacloprid (Table 1). Imidacloprid offered with food induced an increase of total protein content per animal wet weight and an elevation of GST (glutathione S-transferase) activity when juvenile P. scaber were fed with 5 lg/g of imidacloprid (Table 1, Fig. 1a). At 10 and 50 lg/g imidacloprid in food, these two

1330

D. Drobne et al. / Chemosphere 71 (2008) 1326–1334

Table 1 Imidacloprid toxicity to Porcellio scaber in two-week dietary exposure tests, with no-observed (NOEC) and lowest-observed effect concentrations (LOEC) on the basis of concentrations in food Endpoint

Juveniles NOEC

Adults LOEC

NOEL

– 10 50 – – n.a.

>1.95 0.30 0.50 >1.95 >1.95 n.a

(lg/g dry food) Mortality Weight gain Feeding rate Total protein content GST activity Epithelial thickness

>50 5 10 >50a >50a n.a.

LOEL

NOEC

– 0.50 1.95 – – n.a.

>25 >25b <10 >25 10 <10

lg/g b.w./day

LOEC

NOEL

– – 10 – 25 10

>0.32 >0.32b <0.24 >0.32 0.24 <0.24

(lg/g dry food)

LOEL

lg/g b.w./day – – 0.24 – 0.32 0.24

Also included are corresponding effect levels (NOEL and LOEL), based on estimated imidacloprid intake rates (see Fig. 3). n.a. not analysed. a Significant increase (induction) at 5 lg/g imidacloprid in the diet. b Weight loss.

lower weight gain was observed at 10 and 50 lg/g dry food, but feeding rate was reduced only at 50 lg/g of imidacloprid in the food (Table 1). In adults, feeding rate was reduced when animals were fed with 10 lg/g of imidacloprid. In adults we detected decreased GST activity but no effect on total protein content when animals were exposed up to 25 lg/g of imidacloprid (Fig. 1b). Epithelial thickness of the digestive glands was affected already when animals were fed 10 lg/g of imidacloprid (Fig. 2). In this group, the average epithelial thickness of the majority of animals was around 40 lm. Animals with thicker epithelium were almost absent. This trend was even clearer in the animals fed with 25 lg/g of imidacloprid. We therefore consider 10 lg/g imidacloprid in the food as the LOEC for epithelial thinning. Fig. 3 shows the estimated imidacloprid intake rates derived from the food intake rates at the different exposure concentrations in the food. At the same exposure concentration in the diet, adults and juveniles experienced different intake rates because animals consumed different

Fig. 1. Effects of two-week dietary exposure to imidacloprid on glutathione S-transferase (GST) activity in juvenile (a) and adult (b) isopods Porcellio scaber (p < 0.01 – **, p < 0.001 – ***). s – Outlier; Njuveniles (C, 2.5 lg/g, 5 lg/g, 50 lg/g) = 10; Njuveniles (10 lg/g) = 9; Nadults(C) = 25; Nadults (10 lg/g) = 27; Nadults (25 lg/g) = 10.

parameters decreased again, but did not differ from the control. Among the higher-level biomarkers, dose-related

Fig. 2. Effects of two-week dietary exposure to imidacloprid on hepatopancreas epithelial thickness of adult isopods Porcellio scaber. Nadults (C) = 11; s – Outlier; Nadults (10 lg/g) = 12; Nadults (25 lg/g) = 6.

D. Drobne et al. / Chemosphere 71 (2008) 1326–1334

Fig. 3. Imidacloprid intake rates (lg/g body weight/day) of juvenile (a) and adult (b) isopod (Porcellio scaber), estimated from food intake rates during a two-week exposure period (p > 0.05 – *, p > 0.01 – **, p > 0.001 – ***). s – Outlier; Njuveniles (C, 2.5 lg/g, 5 lg/g, 50 lg/g) = 10; Njuveniles (10 lg/ g) = 8; Nadults (C) = 31; Nadults (10 lg/g) = 39; Nadults (25 lg/g) = 21.

amounts of dosed food (Table 1, Fig. 3). The same exposure level also affected juveniles and adults differently. For example, a reduced feeding rate was seen in adults already at 10 lg/g imidacloprid in the diet (equivalent with an intake rate of 0.24 lg/g body weight/day), in juveniles only at 50 lg/g dry food (or 1.95 lg/g b.w./day). 4. Discussion The first of the tested endpoints in juveniles (weight gain) or adults (feeding rate) were affected starting at 10 lg imidacloprid/g food and this concentration therefore is considered the LOEC. This dietary exposure concentration corresponds to an exposure level or intake rate of 0.5 lg/g b.w./day (LOEL) in juveniles, and of 0.24 lg/g b.w./day in adults. It should be noted that data

1331

for adults and juveniles have to be interpreted with caution, since they received food of different quality (leaves versus briquettes), what might have affected the consumption rate irrespective to the amount of added pesticides (Farkas et al., 1996). In addition, bioavailability of imidacloprid may have been different in different food types. Therefore the toxicity data for juveniles and adults can not be compared straightforward. In juveniles reduced weight gain was measured, but in adults net weight loss. Reduction of feeding rate and weight change did not coincide in imidacloprid fed animals. In adults, reduced feeding rate was detected, but weight change did not occur. In juveniles, weight gain was significantly reduced at much lower concentrations of imidacloprid than those affecting feeding rate (Table 1). Also in case of lower-level biomarkers, the response of juveniles and adults differed. In juveniles we detected induced total protein content per animal wet weight and increased GST activity at a low exposure concentration of 5 lg/g dry food, but not at higher treatment levels. In adults however, a decreased GST activity but no changes in total protein content were found. The GSTs are a family of biotransformation enzymes involved in conjugation of a variety of electrophilic metabolites (endogenous and exogenous lipophilic compounds) to glutathione and are most widely studied and most important phase II enzymes. Phase II reactions serve to greatly increase water solubility and, in general, reduce toxicity (Hyne and Maher, 2003). In chemically stressed animals both increased and decreased GST activities are possible, depending on the type of the chemical, time and dose of exposure. No previous data on the effects of imidacloprid on GST activity of isopods are available in the literature. Capowiez et al. (2003) showed no acute effects on GST activities in earthworms exposed up to 1 lg/g imidacloprid in soil, and Jemec et al. (2007a) observed inhibition of GST activity in daphnids exposed to 2.5 mg/l of imidacloprid. It remains to be further investigated whether the evident changes of GST activity in our experiments were a result of reduced fitness or related to detoxification of imidacloprid. In our study a set of biochemical biomarkers and physiological responses was supplemented by histological measures of stress, which is a preferable but not very usual approach in multiple biomarker studies. Histopathological changes are integrators of effects in physiological and biochemical systems in an organism exposed to a natural or anthropogenic stress (Wester et al., 2002). The prime advantage of histopathology is that it provides a window to understanding the functional architecture of cells, tissues, and organs. In isopods, digestive gland epithelial thickness is related to contaminated food (Drobne and Sˇtrus, 1996; Odendaal and Reinecke, 2003). In our study we confirmed the occurrence of epithelial thinning as a result of stress and that reduced feeding rate coincides with reduced epithelial thickness. Toxicity of imidacloprid to P. scaber is in accordance with data reporting higher sensitivity of invertebrates to

1332

D. Drobne et al. / Chemosphere 71 (2008) 1326–1334

Table 2 Toxicity of imidacloprid for different organisms; data from the literature LD50 (mg/kg b.w.) Mammals Birds Bees

450 (oral)a,b 31–238 (oral)?,b 0.14–0.57 (oral)b,c

Other insects Earthworms

0.0027d /

LC50 (mg/kg food or soil) b

1420 (5 days)

LO(A)EL (mg/kg b.w./day)

LO(A)EC (mg/kg)

NO(A)EL (mg/kg b.w./day)

NO(A)EC (mg/kg)

72 (12 d, oral)c 36 (acute oral)a 0.025 (acute oral)b

42 (rats) 69b 2 (7% behavior; 7% reduced visits)c

5.7 (rats)a 3 (acute)b 0.3 (chronic)b 0.012–0.015 (acute oral)b = 0.2 ng/bee 0.010 (chronic)b

– – 0.015 (48 h oral)b

3.48 (soil,7 d) 2.3 (soil, 14 d)e,f

0.2 (soil, 10 d, sperm deformation)e

0.1 (soil, 10 d; sperm deformation)e

Only those reports are listed, which provide toxicity data as concentrations as well as routes and duration of exposure. LD – lethal dose; LC – lethal concentration; LO(A)EL – lowest observed (adverse) effect level; LO(A)EC – lowest observed (adverse) effect concentration; NO(A)EL – no observed (adverse) effect level; NO(A)EC – no observed (adverse) effect concentration. a Tomizawa and Casida (2005). b Anatra-Cordone and Durkin (2005). c Felsot (2001). d Schmuck (2006). e Luo et al. (1999). f Mostert et al., 2002.

imidacloprid when compared with mammals and birds (Table 2). Isopods proved to be less sensitive to imidacloprid than bees and earthworms. However, these data have to be compared with caution, since exposure routes and biomarkers tested in the different invertebrate tests differ. A difficulty with comparing toxicity data on different terrestrial invertebrates is that they are generated in different types of tests with different duration of exposure, exposure levels, and endpoints, resulting in different types of toxicity data. Usually in invertebrates, exposure level is not determined since either consumption rate cannot be recorded or organisms are exposed at the same time via both food and substrate. The test methodology applied in this study proved to be very useful in toxicity testing of chemicals in the terrestrial environment due to its possibility to observe exposure and effects of chemicals at different levels of organization and the ability to determine the actual intake rate of chemicals. Imidacloprid is proposed as an alternative for organophosphorus insecticides. For that reason, here a comparison is made of the potential risk of imidacloprid and the organophosphate diazinon. Diazinon is registered to con-

trol foliage and soil pests of many fruit, nut, vegetable and ornamental crops, and is also used as a veterinary ectoparasiticide (EPA, 2004). The application rates of diazinon for agricultural uses are in the range of 0.5–4 kg of active ingredient/ha depending on the type and method of application (EPA, 2004). The concentration of diazinon in soil can reach up to 10 lg/g soil (Stanek et al., 2006). When comparing effects of imidacloprid (Table 1) and diazinon (Table 3) on the higher-level biomarkers, the LOECs are approximately an order of magnitude higher for diazinon than for imidacloprid (LOECimi = 10 lg/g; LOECdiazinon = 100 lg/g). When effect levels based on chemical intake rates are compared between imidacloprid and diazinon, the lowest observed effect levels (LOELs) are similar (Tables 1 and 3: LOELimi = 0.24 lg/g food; LOELdiazinon = 0.23 lg/g food). To determine the potential risk of imidacloprid and diazinon for isopods, a risk quotient (RQ) is calculated as the ratio estimated exposure levels (PEC) and LOECs: Table 4. From this, imidacloprid seems to be less dangerous (RQ = 0.02) than diazinon (RQ = 1.6).

Table 3 Summary of diazinon toxicity to Porcellio scaber in two-week dietary exposure tests, with no-observed (NOEC) and lowest-observed effect concentrations (LOEC) on the basis of concentrations in food Endpoint

Juveniles NOEC

Mortality Weight gain Induced feeding rate Reduced total protein content Reduced AChE activity

Adults LOEC

NOEL

(lg/g dry food)

lg/g b.w./day

10 P100a P100a 5 <5

0.49 P5.11a P5.11a 0.23 <0.23

50 – – 10 5

LOEL

NOEC

2.71 – – 0.49 0.23

50 >100b P100a 50 10

LOEC

NOEL

100 – – 100 50

0.93 >2.54b P2.54a 0.93 0.25

(lg/g dry food)

LOEL

lg/g b.w./day 2.54 – – 2.54 0.93

Also included are corresponding effect levels (NOEL and LOEL), based on estimated diazinon intake rates (adapted from Stanek, 2004, and Stanek et al., 2006). a Decreased but not statistically significant compared to control. b Weight loss.

D. Drobne et al. / Chemosphere 71 (2008) 1326–1334 Table 4 Risk assessment of imidacloprid and diazinon for isopods (Porcellio scaber) using the results of this study and the studies of Stanek (2004) and Stanek et al. (2006) (see Table 3) Imidacloprid PEC (lg/g) Estimated daily dosec PEL/day (lg/g wet wt./day) LOAEC isopodsd (lg/g food) LOAEL isopodsd (lg/g wet wt./day) NOEL isopodsd (lg/g wet wt./day) NOAEL isopodsd (lg/g wet wt./day) RQ = PEC/LOAEC RfD-E = NOEL/100e (lg/g wet wt./day) RfD-A = NOAEL/100e (lg/g wet wt./day) HQ: estimated daily dose/RfD-A

a

Diazinon

0.1–0.2 0.008

7–80b 3.2

10 0.24 0.24 <0.24 0.02 0.0024 <0.0024 >3.33

50 2.54 <0.23 0.49 1.6 <0.0023 0.0049 653

Predicted environmental concentrations (PEC) and exposure levels (PEL) are estimated from normal recommended dosages. LO(A)EC, LO(A)EL and NO(A)EL refer to lowest or no-observable (adverse) effects concentrations or levels, with adverse being used to indicate reduction of higherlevel endpoints (survival, weight gain, feeding rate); NOEL also includes lower-level biomarker, like reduction of acetyl cholinesterase (diazinon) or glutathione S-transferase (imidacloprid) activity. RQ = risk quotient; RfD-E = reference dose vased on effects on all endpoints, RfD = reference dose based on adverse effects on higher-level endpoints. a Anatra-Cordone and Durkin (2005). b EPA (2004). c Assuming consumption rate of 0.04 g/g wt. b.w./day. d Juveniles and adults. e Assessment factor of 100, including two factors of 10 for interspecies and intraspecies variability, respectively.

Since data on exposure levels usually are lacking, calculating acute or chronic reference doses (RfDs) is no common practice in environmental risk characterization. On the basis of the toxicity tests with P. scaber described in this study, acute RfDs for both imidacloprid and diazinon were calculated (Table 4). We calculated RfDs from NOELs taking into account all endpoints measured, but also from no observed adverse effect levels (NOAELs) that include only the higher-level endpoints like survival, weight change and feeding rate. Hazard quotients (HQ) were calculated applying a safety factor of 100 to NO(A)ELs to derive a reference dose (RfD). HQ, defined as the estimated daily dose divided by the RfD, is much higher for diazinon (HQ = 653) than for imidacloprid (HQ = >3.33) (Table 4). This suggests that isopods are more sensitive to imidacloprid than to diazinon. Since there are not enough data on environmentally relevant concentrations of both pesticides, both RQ and HQ values have to be considered of limited value. When NOEL is determined on the basis of a specific biomarker, such as AChE activity in the case of organophosphorus pesticide diazinon, the RfD values are very low. However, for determining effects of imidacloprid no such specific target sites were evaluated, but rather more general biomarkers, which change as a result of imidacloprid’s action on its primary receptor nAChR. Consequently RfD values for imidacloprid are higher, and are interpreted as lower risk. Hu et al. (2000) reported comparative toxic-

1333

ity data for mortality, with diazinon being less toxic than imidacloprid for Rhagoletis pomonella, (Diptera, Insecta) after ingestion and tarsal contact. An aspect not included in this comparative risk assessment is the stability or environmental persistence of the two chemicals. Imidacloprid is more persistent in soil, with half-lifes (DT50s) for degradation in soil ranging from 27 to 230 days, while diazinon is degraded much faster with DT50s of 5–90 days (Fossen, 2006; see also Jemec et al., 2007a for a review). As a consequence, there is a much greater risk for chronic exposure to imidacloprid than for diazinon. Our comparative toxicity study stresses the importance of careful selection of biomarkers when risk values are calculated. Perhaps it would make sense to also consider toxicity data obtained in vertebrate toxicity studies to support the selection of the most suitable biomarkers. This is of particular importance when toxicity data are decisive for continued use or withdrawal of a particular pesticide. Our results indicate high toxicity of imidacloprid to the non-target terrestrial arthropod P. scaber when compared to birds and mammals. Imidacloprid therefore is not specific only to insects, but also affects other invertebrates. Our study stresses the importance of carefully selecting biomarkers that are useful for determining NOELs. Specific biomarkers might overestimate the risk. Biomarkers at a higher-level of biological complexity that integrate lowerlevel responses perhaps are the best choice for determining the NO(A)EC. In conclusion, if future environmental levels of imidacloprid remain as low as currently determined, there is little risk of imidacloprid to isopods and other non-target invertebrates. When the use of imidacloprid will further increase, this pesticide may however, pose a serious risk to non-target organisms. Since invertebrates play a crucial role in the ecosystem (Wilson, 1987), there is a need for more toxicity data on invertebrates to characterize risk of imidacloprid in environment. Acknowledgments The research was funded by the Ministry of Education, Science and Sport of the Government of the Republic of Slovenia, through Grant No. J1-3186. References Anatra-Cordone, M., Durkin, P., 2005. Imidacloprid. Human health assessment and ecological risk assessment – Final report. Syracuse Environmental Research Associates, Inc., New York, SERA TR 0543-24-03a, December 28, 2005. APVMA, 2003. The reconsideration of registrations of products containing diazinon and their labels. Part 1: product cancellations. Australian Pesticides & Veterinary Medicines Authority. Bayer Technical Information, ConfidorÒ, 2000. Bayer, Germany. Bradford, M.M., 1976. Rapid and sensitive method for quantitation of microgram quantities of protein utilizing principle of protein–dye binding. Anal. Biochem. 72, 248–254.

1334

D. Drobne et al. / Chemosphere 71 (2008) 1326–1334

Buckingham, S.D., Lapied, B., LeCorronc, H., Grolleau, F., Sattelle, D.B., 1997. Imidacloprid actions on insect neuronal acetylcholine receptors. J. Exp. Biol. 200, 2685–2692. Capowiez, Y., Rault, M., Mazzia, C., Belzunces, L., 2003. Earthworm behaviour as a biomarker – a case study using imidacloprid. Pedobiologia 47, 542–547. Cobb, G.P., Mellott, R., Brewer, W., Bens, C.M., Kendall, R.J., 2000. Diazinon dissipation from vegetation, occurrence in earthworms, and presence in avian gastrointestinal tracts collected from apple orchards following D-Z-N 50 W application. Environ. Toxicol. Chem. 19, 1360– 1367. Cox, C., 2001. Insecticide factsheet/imidacloprid, Northwest Coalition for Alternatives to Pesticides (NCAP). J. Pest. Ref. 21, 15–21. Drobne, D., Sˇtrus, J., 1996. The effect of Zn on the digestive gland epithelium of Porcellio scaber (Isopoda, Crustacea). Pflu¨gers. Arch. 431, 247–248. Drobne, D., 1997. Terrestrial isopods – a good choice for toxicity testing of pollutants in the terrestrial environment. Environ. Toxicol. Chem. 16, 1159–1164. Drobne, D., Drobne, S., 2005. Application of computer microscopy for histopathology in isopod toxicity studies. In: Ostrander, Gary Kent (Ed.), . In: Techniques in Aquatic Toxicology, vol. 2. Taylor & Francis, Boca Raton, Florida, USA, pp. 137–146 (Chapter 7). EPA, 2004. Interim registration eligibility decision. Prevention pesticides and toxic substances (7508C). 738-R-04-006. . Farkas, S., Hornung, E., Fischer, E., 1996. Toxicity of copper to Porcellio scaber Latr. (Isopoda) under different nutritional conditions. Bull. Environ. Contam. Toxicol. 57, 582–588. Felsot, 2001. Admiring risk reduction. Does imidacloprid have what it takes? Agrichem. Environ. News 186, 2–13. Felsot, A.S., Ruppert, J.R., 2002. Imidacloprid residues in Wilapa Bay (Washington State) water and sediment following application for control of burrowing shrimp. J. Agric. Food Chem. 50, 4417–4423. Fossen, M., 2006. Environmental fate of imidacloprid. California Department of Pesticide Regulation. . Habig, W.H., Pabst, M.J., Jakoby, W.B., 1974. Glutathione S-transferases – first enzymatic step in mercapturic acid formation. J. Biol. Chem. 249, 7130–7139. Hu, X.P., Prokopy, R.J., Clark, J.M., 2000. Toxicity and residual effectiveness of insecticides on insecticide-treated spheres for controlling females of Rhagoletis pomonella (Diptera: Tephritidae). J. Econ. Entomol. 93, 403–411. Hyne, R.V., Maher, W.A., 2003. Invertebrate biomarkers: links to toxicosis that predict population decline. Ecotoxicol. Environ. Safety 54, 366–374. Jemec, A., Tisler, T., Drobne, D., Sepcic, K., Fournier, D., Trebse, P., 2007a. Comparative toxicity of imidacloprid, of its commercial liquid formulation and of diazinon to a non-target arthropod, the microcrustacean Daphnia magna. Chemosphere 68, 1408–1418. Jemec, A., Drobne, D., Tisler, T., Trebse, P., Ros, M., Sepcic, K., 2007b. The applicability of acetylcholinesterase and glutathione S-transferase in Daphnia magna toxicity test. Comp. Biochem. Physiol. C–Toxicol. Pharmacol. 144, 303–309. Kammenga, J.E., Dallinger, R., Donker, M.H., Ko¨hler, H.-R., Simonsen, V., Triebskorn, R., Weeks, J.M., 2000. Biomarkers in terrestrial invertebrates for ecotoxicological soil risk assessment. Rev. Environ. Contam. Toxicol. 164, 93–147. Kaschl, U., Zidar, P., Sˇtrus, J., Storch, V., 2002. Food choice experiments with cadmium nitrate dosed food in terrestrial isopod Oniscus asellus (Crustacea). Acta Biol. Slov. 45, 3–14. Lapanje, A., Rupnik, M., Drobne, D., 2007. Gut bacterial community structure (Porcellio scaber, Isopoda, Crustacea) as a measure of

community level response to long-term and short-term metal pollution. Environ. Toxicol. Chem. 26, 755–763. Lind, R.J., Hardick, D.J., Blagbrough, I.S., Potter, B.V.L., Wolstenholme, A.J., Davies, A.R.L., Clough, M.S., Earley, F.G.P., Reynolds, S.E., Wonnacott, S., 2001. [H-3]-Methyllycaconitine: a high affinity radioligand that labels invertebrate nicotinic acetylcholine receptors. Insect Biochem. Mol. Biol. 31, 533–542. Lindstrom, J., Anand, R., Peng, X., Gerzanich, V., Wang, F., Li, Y. B., 1995. Neuronal nicotinic receptor subtypes. In: Functional Diversity of Interacting Receptors. Ann. NY Acad. Sci. 757, 100–116. Liu, H.X., Song, J.M., Zhang, S.S., Qu, L.B., Zhao, Y.F., Wu, Y.J., Liu, H.M., 2005. Analysis of residues of imidacloprid in tobacco by highperformance liquid chromatography with liquid–liquid partition cleanup. Pest Manage. Sci. 61, 511–514. Løkke, H., Van Gestel, C.A.M. (Eds.), 1998. Handbook of Soil Invertebrate Toxicity Tests. John Wiley & Sons, Chichester, New York, USA. Luo, Y., Zang, Y., Zhong, Y., Kong, Z., 1999. Toxicological study of two novel pesticides on earthworm Eisenia foetida. Chemosphere 39, 2347– 2356. Matsuda, K., Buckingham, S.D., Kleier, D., Rauh, J.J., Grauso, M., Sattelle, D.B., 2001. Neonicotinoids: insecticides acting on insect nicotinic acetylcholine receptors. Trends Pharmacol. Sci. 22, 573–580. Mostert, M.A., Schoeman, A.S., van der Merwe, M., 2002. The relative toxicities of insecticides to earthworms of the Pheretima group (Oligochaeta). Pest Manage. Sci. 58, 446–450. Odendaal, J.P., Reinecke, A.J., 2003. Quantifying histopathological alterations in the hepatopancreas of the woodlouse Porcellio laevis (Isopoda) as a biomarker of cadmium exposure. Ecotoxicol. Environ. Safety 56, 319–325. Roberts, A.P., Oris, J.T., 2004. Multiple biomarker response in rainbow trout during exposure to hexavalent chromium. Com. Biochem. Physiol. 138C, 221–228. Schmuck, R., 2006. No causal relationship between GauchoÒ seed dressing in sunflowers and the French bee syndrome. PflanzenschutzNachrichten Bayer 52/99 (3), 257–299. Stanek, K., 2004. Sensitivity and specificity of acetylcholinesterase activity in Porcellio scaber (Isopoda, Crustacea) as a biomarker of diazinon exposure. Master’s thesis, Nova Gorica. Polytechnic, School of Environmental Sciences. Nova Gorica, Slovenia. Stanek, K., Drobne, D., Trebse, P., 2006. Linkage of biomarkers along levels of biological complexity in juvenile and adult diazinon fed terrestrial isopod (Porcellio scaber, Isopoda, Crustacea). Chemosphere 64, 1745–1752. TDC Environmental, 2003. Insecticide market trends and potential water quality implications. TDC Environmental, LLC, San Mateo, CA, USA. Tomizawa, M., Casida, J.E., 2005. Neonicotinoid insecticide toxicology: mechanisms of selective action. Ann. Rev. Pharmacol. Toxicol. 45, 247–268. Ware, G.W., Whitacre, D.M., 2004. An Introduction to Insecticides, fourth ed., MeisterPro Information Resources, A division of Meister Media Worldwide, Willoughby, Ohio, USA. . Wester, P.W., van der Ven, L.T.M., Vethaak, A.D., Grinwis, G.C.M., Vos, J.G., 2002. Aquatic toxicology: opportunities for enhancement through histopathology. Environ. Toxicol. Pharmacol. 11, 289–295. Wilson, E.O., 1987. The little things that run the World (the importance and conservation of invertebrates). Conserv. Biol. 1, 344–346. Zidar, P., Bozic, J., Strus, J., 2005. Behavioral response in the terrestrial isopod Porcellio scaber (Crustacea) offered a choice of uncontaminated and cadmium-contaminated food. Ecotoxicology 14, 493–502.