Biomarkers for the assessment of chlorpyrifos effects on earthworms and on soil functional parameters

The objective of this work was to evaluate the effects of chlorpyrifos on earthworms and on soil functional parameters. An integrated laboratory-fi eld study was performed in a wheat fi eld in Argentina, sprayed with chlorpyrifos at two recommended application rates (240 or 960 g ha a.i.). Laboratory tests included neutral red retention time, comet assay (single cell gel electrophoresis), and avoidance behavior, each using the earthworm Eisenia andrei exposed in soil collected 1 or 14 days after pesticide application, and the bait-lamina test. Field tests assessed organic matter breakdown using the litterbag and bait-lamina assays. Earthworm populations in the fi eld were assessed using formalin application and hand-sorting. The neutral red retention time and comet assays were sensitive biomarkers to the effects of chlorpyrifos on the earthworm E. andrei; however, the earthworm avoidance test was not suffi ciently robust to assess these effects. Feeding activity of soil biota, assessed by the bait lamina test, was signifi cantly inhibited by chlorpyrifos after 97 days, but recovered by the 118 day of the test. Litterbag test showed no signifi cant differences in comparison to controls. Earthworm abundance in the fi eld was too low to adequately test the sensitivity of this assessment endpoint.


Introduction
It is increasingly recognized that the protection of soils and their inherent communities must become a primary goal of environmental policy worldwide (Römbke et al., 2005;Filser et al., 2008).
Chlorpyrifos (O,O-diethyl-O-3,5,6-trichloro-2-pyridyl phosphorothioate), a broad spectrum systemic organophosphorus anticholinesterase insecticide, is widely used in Argentina in direct soil application for the control of agricultural pests.Pesq. agropec. bras., Brasília, v.44, n.8, p.874-880, ago. 2009 The toxicity of pesticides to soil organisms depends on the compound bioavailability, which is affected by the physicochemical properties of the compound and the soil, and by the uptake routes of exposed organisms.Therefore, ecotoxicity studies can benefi t from using experimental designs that improve data relevance for local exposure conditions in the fi eld (Yu et al., 2006;Filser et al., 2008).However, due to this complexity, assessment of the risks for soil organisms cannot rely exclusively on chemical analysis.In-vivo assays using sentinel species provide a more reliable assessment of toxicity because they are more representative of the natural soil conditions.
Earthworms are often used as terrestrial model organisms for ecotoxicity testing, because of their importance for the structure and function of soil ecosystems.Standard methods with the compost earthworms Eisenia fetida and E. andrei have been used in ecotoxicological research for decades (Organization for Economic Cooperation and Development, 1984).Biomarkers on sentinel organisms address the questions of bioavailability by only responding to the bioactive fraction.They have been used extensively to document and quantify both the exposure to and the effects of environmental pollutants, and have the potential to rapidly recognize effects that may affect populations and communities.
The types of biological responses that could be considered as biomarkers range from the molecular, subcellular and cellular to the effects on the intact organism, the population, or community structure, and perhaps also the structure and function of ecosystems (Colgan et al., 2003).Nonspecifi c suborganism biomarkers, such as lysosomal membrane stability measured as neutral red retention time, and DNA damage estimated by the comet assay are indicative of effects of more than one detrimental factor, and because of their sensitivity they may be relevant to assessing the risk of multiple chemicals (Xiao et al., 2006;Reinecke & Reinecke, 2007).In order to validate suborganism biomarkers, responses occurring at different levels of biological organization should be addressed (Colgan et al., 2003;Vasseur & Cossu-Leguille, 2003).
Considerable attention has been given to the effects of agricultural practices on soil organisms, especially earthworms.The abundance and activity of earthworms in arable lands depends strongly on management practices; therefore, earthworms can act as potential bioindicators of land use practices (Suthar, 2009).Ecological tests, such as avoidance behaviour, based on the earthworm's ability to detect a toxicant and move away, may be sensitive and fast (Hund-Rinke et al., 2003).Ecotoxicological information gained from the earthworm studies can be supplemented by a more holistic (i.e., representative of multiple levels of ecosystem organization) endpoint, such as organic matter decomposition, which can be assessed in either laboratory or fi eld studies (Knacker et al., 2003).
The objective of this work was to evaluate the effects of chlorpyrifos on earthworms and on soil functional parameters.

Materials and Methods
An integrated laboratory-fi eld study, on a wheat fi eld in Argentina, was performed.Wheat was machine-planted in a fi eld in Santa Fe Province, Argentina (33°55'24''S, 61°16'42''W), on a clay loam soil with an abundant residue cover.The mean annual rainfall and temperature are 726 mm and 16.5°C, respectively.The experimental design consisted of randomized complete blocks with three replicates, each block comprising three plots of 13x5 m.The distance between blocks was at least 200 m.At each block, randomly chosen plots were sprayed with commercial chlorpyrifos 240 g ha -1 a.i.(low rate) or 960 g ha -1 a.i.(high rate), 30 days after seeding, under controlled conditions, using a calibrated hand-pressure sprayer.Control plots were sprayed with water.The study was conducted in winter/spring (July 2005-November 2005).
Bioassays were performed using adult organisms of E. andrei, with a well developed clitellum, which were Pesq. agropec. bras., Brasília, v.44, n.8, p.874-880, ago. 2009 maintained in laboratory culture, by exposure to soil samples sieved (≤2 mm) and adjusted to 50-60% of the soil water holding capacity (WHC = 48.6 -+2.4%).Unless other specifi cations, OECD/ISO guidelines were followed for the tests.To perform cellular/ subcellular assays, six adult earthworms were added to 300 g soil (three replicate containers per treatment).After a 7-day exposure, each earthworm was washed with water, dry blotted using a fi lter paper, and placed individually in a glass conical tube containing 1.5 mL of phosphate-buffered saline solution (PBS) (Ca ++ , Mg 2+ free).Worms were subjected to three short bouts of electric stimulation (<1s) at 6 V DC.The coelomic fl uid was immediately transferred to an Eppendorf tube.Viability was determined in an improved Neubauer hemocytometer, by diluting 10 μL of cell suspension with 10 μL of 0.4% trypan blue in PBS.The percentage viability was calculated based on the percentage of unstained cells.
The neutral red retention assay on the coelomocyte cells of each worm was measured according to Weeks & Svendsen (1996) with slight modifi cations.Twenty microliters of coelomic fl uid were placed on a microscope slide, and the cells were allowed to adhere for 60 s, prior to the application of 20 μL of neutral red working solution (80 μg mL -1 ) and a coverslip.Each slide was scanned for 2 min, at 5-min intervals under a light microscope (400 x).Observation went on until the ratio of cells with fully stained cytoplasm exceeded 50% of the total number of cells counted.This time was recorded as the neutral red retention time (NRRT).
The comet assay was performed on the same coelomic fl uid (Casabé et al., 2007).We mixed 10 μL of the coelomocyte suspension with 75 μL of 0.75% low-melting-point agarose at 37ºC, and spread it over a microscope slide precoated with 100 μL 1% normal-melting-point agarose.After 5 min of solidifi cation at 4ºC, a second layer of low-melting-point agarose was placed on top and left to harden for 5 min at 4ºC.Slides were placed in alkaline lysis solution (2.5 mol L -1 NaCl, 100 mmol L -1 Na 2 EDTA, 10 mmol L -1 Tris, 1% N-lauryl sarcosinate, 10% DMSO, and 1% Triton X-100, pH 10) and stored overnight at 4ºC.After lysis, slides were rinsed with neutralization buffer (0.4 mol L -1 Tris-HCl, pH 7.5) and immersed in freshly prepared alkaline electrophoresis solution (300 mmol L -1 NaOH and 1 mmol L -1 Na 2 EDTA) at 4ºC for 12 min, to allow DNA unwinding.Electrophoresis was conducted for 20 min at 25 V (1 V cm -1 ) and a starting current of 250 mA.One hundred cell nuclei per slide were rated visually and classifi ed into four categories, according to the tail intensity (size and shape).The extent of DNA damage was measured as damage index (DI), a weighted value of damage according to cells distribution, defi ned as DI = ∑(ni x i), in which ni is the number of cells with damage class i (0, 1, 2 or 3).
The avoidance behaviour was studied according to ISO 17512 (International Organization for Standardization, 2008) with slight modifi cations.Plastic, rectangular, two-chamber containers (20x10x10 cm) were used.Containers were divided into half with a plastic split, and 520 g of control soil were placed in one half of the container, and 520 g of treated soil was placed on the other side.After the split was removed, ten adult earthworms were placed on the centreline in the soil surface, and the containers were covered allowing suffi cient aeration.After three days of incubation, the split was reintroduced, and the number of individuals in each compartment was counted.Four replicates were used for each test.For each block, a control dual test was performed with four replicates.
The laboratory bait-lamina test was performed as an adaptation of Helling et al. (1998).Containers were used with 350-400 g of soil, four bait-laminas and six earthworms, in each of the four replicates per treatment per block.The bait-lamina is a plastic strip perforated at 5-mm distances with 16 small (1-mm diameter) holes, fi lled with cellulose, wheat bran and activated carbon.After exposure for three days, the number of pierced holes in each lamina was counted.
Field assays to determine the effects of chlorpyrifos on organic matter decomposition included the litterbag test and the bait-lamina test.The litterbag test was performed according to Ingelsfi eld (1989) using bags fi lled with lucerne (Medicago sativa L.), with mesh size of either 0.2 mm (microfl ora, microfauna and mesofauna allowed), or 3.6 mm (microfl ora, meso-and macrofauna allowed).Litterbags were buried at 10 cm depth in the soil, before treatments with the pesticides (two replicates per treatment per block).Litterbags were sampled after 97 and 118 days.The contents were dried (50ºC, 48 hours) and weighed.Organic matter breakdown was calculated as the percentage of organic material mass loss at the sampling Pesq.agropec.bras., Brasília, v.44, n.8, p.874-880, ago. 2009 time t [(start weight-end weight) x 100/start weight].Bait-lamina test was performed according to Von Törne (1990).Sixteen bait-lamina strips (as a 4x4 matrix) were inserted vertically into the top soil layer (three replicates per treatment per block).After 14, 97, and 118 days of exposure, laminas were removed from the soil and examined.The number of pierced holes in each lamina was counted.
To study natural earthworm populations, formalin (0.2%) was used to extract them from the soil.Aclitellate and clitellate earthworms were counted and taxonomically identifi ed.
Statistical analyses were performed with GraphPad InStat 3 (GraphPad Software, San Diego, USA

Results and Discussion
Soils treated with 240 g ha -1 a.i.chlorpyrifos showed the same average concentration (0.02±0.01 mg kg -1 ) after 1 or 14 days of treatment application.In plots sprayed with 960 g ha -1 a.i., concentration was 0.10±0.02mg kg -1 after 1 day, and 0.03±0.01mg kg -1 after 14 days from treatment application.
Chlorpyrifos was neither detected in soils sampled 97 days after treatment, nor in the control plots (detection limit: 0.01 mg kg -1 ).As chlorpyrifos soil concentration was similar for both treatments after 14 days, and due to the large number of soil samples to be processed, laboratory bioassays were only performed on samples of soils sprayed with the high rate of pesticide.Lysosomal neutral red retention time was signifi cantly reduced in earthworms exposed for seven days to chlorpyrifos low-rate and high-rate treated soils, collected one day after spraying, when compared to controls (Figure 1).A signifi cant concentration-response relationship was observed (r = 0.888, p<0.001).The reduction was also observed in earthworms exposed to soils sampled 14 days after treatment.
Neutral red retention assay has been used to assess concentration-response relationships for a variety of metals and organic compounds (Svendsen et al., 2004;Gastaldi et al., 2007).Few studies, however, showed effects on neutral red retention time in earthworms exposed to chlorpyrifos contaminated soils (Hodge et al., 2000;Casabé et al., 2007;Reinecke & Reinecke, 2007).Our results showed that the neutral red retention time was sensitive to the low chlorpyrifos concentrations present in soil, and may constitute an early indication of impending physiological damage in the earthworms.
Chlorpyrifos produced a signifi cant increase in damage index with respect to controls in coelomocytes of earthworms exposed to soils sampled one day after treatments (Figure 2).A signifi cant increase was also observed on high-rate treated soils collected 14 days after spraying, although DNA migration was lower.In all experiments, the viability of cells was approximately 95%.Damage to DNA, as measured with comet assay, may lead to mutations, strand breaks, altered bases and, fi nally, health disorders, resulting in severe disturbances in ecosystems which may lead, in some cases, to an elevated extinction risk of sensitive species.
Although comet assay is considered an important biomarker of DNA damage in earthworm ecotoxicology (Fourie et al., 2007), few studies have been reported on terrestrial environments (Martin et al., 2005;Xiao  Pesq.agropec.bras., Brasília, v.44, n.8, p.874-880, ago. 2009Brasília, v.44, n.8, p.874-880, ago. et al., 2006)).The comet assay applied on earthworm coelomocytes showed a high level of DNA damage, by exposure to chlorpyrifos treated soils.Neutral red retention and comet assays revealed alterations at subcellular level, and can be regarded as indicators to be used in the assessment of soil earthworms health.Previous results on a soybean fi eld of the Santa Fe Province, sprayed with chlorpyrifos, showed a similar pattern in both cellular biomarkers (Casabé et al., 2007).In terms of the avoidance, earthworms exposed to soils sampled 14 days after treatment had varied results: when exposed to soils from block 1, earthworms preferred the treated substrate; in contrast, worms tended to avoid treated soils from blocks 2 and 3 (Figure 3).The responses were statistically signifi cant.Soil physicochemical analysis showed no differences between blocks.No signifi cant difference was found in the distribution of the worms between both chambers of the containers in dual control tests, which indicates an even distribution of individuals among the two sections of the test container.
Avoidance behaviour is an ecologically relevant endpoint, directly related to the energy budget of the worms, and indirectly to the soil structure.Exposure to pesticides that alters earthworm behaviour can induce migratory behaviour, which can lead to modifi cations in population abundance or biomass, and to changes in species diversity.Although avoidance tests with earthworms have been considered as suitable screening tools for the assessment of potentially contaminated soils (da Luz et al., 2004;Sousa et al., 2008), few studies were done with chlorpyrifos treated soils (Hodge et al., 2000;Zhou et al., 2007).The results of this study showed that the avoidance behaviour was not a suffi ciently sensitive endpoint for assessment of the effects of chlorpyrifos on the earthworms.
Bait-lamina laboratory test showed a nonsignifi cant decrease in the bait consumption of earthworms exposed to treated soils collected 1 and 14 days after spraying, compared to controls (Figure 4).These results contrast with those reported previously for a soybean fi eld on a clay silty soil sprayed with chlorpyrifos (Casabé et al., 2007).Physical structure and chemistry of the soils have a strong infl uence on this functional endpoint, and could account for the observed differences.In fi eld experiments, feeding activity of soil macrofauna assessed with the fi eld bait-lamina test increased with increasing exposure time (Figure 5).Fourteen days after introduction of the bait-laminas into soils, the substrate consumption rates, measured as the percentage of open holes in the bait-laminas, were very low, probably as a consequence of the low density of earthworms and of the short exposure time.No signifi cant differences between treated and control soils were observed.Feeding activity was signifi cantly decreased in both treatments, compared to control soils, after 97 days of exposure; however, no concentration-dependent relationship was observed.After 118 days exposure, a nonsignifi cant decrease was observed.
Chlorpyrifos treatment of the soil did not affect litter decomposition, after 97 or 118 days of the litterbag exposure in the fi eld (Figure 6).The overall litter  decomposition rate was very slow.No difference was detected between the decomposition rates of the leaves enclosed in the fi ne and coarse mesh litterbags.
Decomposition of organic matter is one of the most integrating processes in the soil ecosystems.Pesticides affecting this function might adversely infl uence nutrient cycling and soil fertility (Förster et al., 2004).The two assayed fi eld methods showed a different pattern of response that could be attributed to the different quality and way of exposure of the organic matter used (Knacker et al., 2003).Besides, both functional endpoints are closely related to soil moisture content and abundance of earthworms in soil (Förster et al., 2004).The number of earthworms was very low in the fi rst 14 days of the experiment -Octolasion cyaneum (Lumbricidae) and Microscolex dubius (Acanthodrilidae) ocurrences -, and no signifi cant differences between control and treated soils were observed.No earthworms were found at the subsequent sampling dates.

Conclusions
1.The neutral red retention time and comet assays are sensitive biomarkers of chlorpyrifos effects, on the exposed earthworm Eisenia andrei.
2. The earthworm avoidance test was not suffi ciently robust for assessing the effects of chlorpyrifos on Eisenia andrei.
3. Feeding activity of soil biota, assessed by the bait-lamina test, is signifi cantly inhibited by chlorpyrifos after 97 days, but recovers by the 118 th day of the test.
4. Chlorpyrifos treatment of the soil does not affect litter decomposition after the 118-day exposure of the litterbag in the fi eld.
5. The earthworm abundance was too low to adequately test the sensitivity of this assessment endpoint.

Figure 1 .
Figure 1.Neutral red retention time, in coelomocytes of Eisenia andrei exposed for seven days to control and chlorpyrifos-treated soils, sampled at different times after treatments.Mean±SD (three earthworms per treatment per block).Different letters denote signifi cant differences between groups.

Figure 2 .
Figure 2. Effects on DNA migration of coelomocytes of E. andrei exposed to control and chlorpyrifos-treated soils, sampled at different times after treatments in the comet assay.Mean±SD (three earthworms per treatment per block).*Signifi cantly different from control, at 0.1% of probability.

Figure 3 .
Figure 3. Distribution of Eisenia andrei on each section of the dual chamber test container, in the avoidance behavior assay.Mean±SD (four replicates per assay).*Signifi cantly different from control, at 1% of probability.

Figure 4 .
Figure 4. Feeding activity (three days) of Eisenia andrei in controls and chlorpyrifos-treated soils, sampled at different times after treatments in the bait-lamina laboratory assay.Mean±SD (four replicates per treatment per block).

Figure 5 .
Figure 5. Feeding activity of soil fauna, in controls and chlorpyrifos-treated soils in the bait-lamina fi eld assay.Mean±SD (three replicates per treatment per block).*Signifi cantly different from control, at 0.1% of probability.