Integrated evaluation of the biological response of the earthworm Eisenia fetida using two glyphosate exposure strategies: soil enriched and soils collected from crops in Southeastern Mexico

Under laboratory conditions, the toxicological effects of pesticides tend to be less variable and realistic than in �eld studies, limiting their usefulness in environmental risk assessment. In the current study, the earthworm Eisenia fetida was selected as a bioindicator for assessing glyphosate toxic effects in two different trials to solve this question. In Trial 1, the worms were exposed for 7 and 14 days to concentrations of a commercial glyphosate formulation (1 to 500 mg a.i. kg − 1) currently in the �eld. For Trial 2, the worms were kept in nine soils collected from different plots with crops for 14 days of exposure. In both experiments, glutathione S-transferase (GST), superoxide dismutase (SOD), catalase (CAT), glutathione peroxidase (GPx), and acetylcholinesterase (AChE) activities and contents of lipid peroxidation (LPO) were evaluated. In T1, glyphosate formulation produced a 40% inhibition of AChE activity and a signi�cant increase in GST, SOD, CAT, GPx activities, and LPO contents of E. fetida on day 7. In T2, higher concentrations of glyphosate were detected in soils of soybean, papaya, and corn (0.92, 0.87, and 0.85 mg kg − 1), which induced a positive correlation between the levels of glyphosate residues with GST, SOD, CAT, GPx, and LPO, and negative with AChE. These �ndings indicate that crop soils polluted with glyphosate elicited higher oxidative stress than in laboratory conditions, con�rmed by IBRv2, PCA, and AHC analysis.


Introduction
The expansion of modern and industrialized agriculture in tropical regions is highly dependent on the extensive use of modern pesticides, fertilizers, genetically modi ed crops, and other technologies that increase the risk of negative impacts on the physiology of non-target organisms and the deterioration of the structure and function of soil ecosystems (De Silva 2009).
Glyphosate [GLY; N-(phosphonomethyl)-glycine] is a non-selective, systemic, broad-spectrum organophosphate herbicide used to control most annual and perennial weeds by disrupting the synthesis of aromatic amino acids essential for plant growth (Benbrook 2016).Currently, it is the best-selling herbicide in the world (60% of the world market).It has several commercial formulations distributed in concentrations of 350 to 720 grams of active ingredient (a.i) per L or kg-1.In these concentrations, the active ingredient's penetration through the treated plants' waxy surfaces is favored increasing its.
However, commercial formulations contain surfactants that have been documented to be much more toxic than pure GLY (Mertens et al. 2017).
In Mexico, around 15,785 tons of herbicides are applied to agricultural soils each; of these herbicides, GLY (27%) is one of the most used due to formulations and economical price (Pastrana-Cervantes et al.

2017
).When GLY is applied to foliar, a signi cant amount of this herbicide can reach the soil or be released through the plant metabolism, where it readily binds to soil clay minerals and essential metal ions.In the soils, GLY binding to clays through its phosphonic acid fraction favored by its phosphate sorption rate (pKa1 = 2.27, pKa2 = 5.58, pKa3 = 10.25), and could reach a maximum concentration of 0.45 mg kg-1 when the application rate is about 1 kg ha-1.These interactions overestimate the actual level of GLY in the upper part of the soil (0.5 to 4.3 mg kg-1).In addition, GLY can form chelates and, in the soils, remain between 2-215 days (Peruzzo et al. 2008; Gomes et al. 2014;Maqueda et al. 2017).For example, GLY concentrations in agricultural soils in the European Union have been documented at mean values ranging from 0.05 to 1.14 mg kg-1 (Silva et al. 2018).
The release of GLY and its chronic availability is susceptible to polluting the soils affecting their quality and health, and can signi cantly damage non-target organisms such as earthworms (Mesnage et al. 2020).Therefore, earthworms living in these soils suffer the toxic effects of sublethal exposure (Givaudan et al. 2014).
Acute ecotoxicological tests are fundamental tools to evaluate the toxicity of pesticides in soil organisms and are considered a preliminary step (lower level) in environmental risk assessments (Alves et al. 2013; Daam et al. 2019).However, such tests are complementary (quick and sensitive results) because they do not re ect ecological realism (changes in temperature, climatic events, particular physicochemical properties of the type of soil, synergism of contaminants).In addition, in the environment, crop soils could slowly become polluted over time (long term), and also can occur adaptability of organisms affecting, in this way, the multiple levels of ecosystem organization.To address these concerns, some authors have conducted studies of the exposure of earthworms to eld-collected soils (natural soils) as test substrates.In addition, other tests with arti cial sediments proposed by the OECD to simulate tropical arti cial soils (TAS) are proposed, which include experimentation at higher temperatures and local species from the tropics following the guidelines of the International Organization for Standardization (ISO) (Jänsch et al. 2006; Kuperman et al. 2006; García-Santos and Keller-Forrer 2011; Niemeyer et al. 2017; Bandeira et al. 2020).It has been documented that GLY and its surfactants can affect earthworm physiological and biochemical parameters under controlled laboratory conditions.Sublethal effects include the production of reactive oxygen species (ROS), and lipid peroxidation (LPO), increased activity of the antioxidant defense system, including catalase (CAT) and glutathione S-transferase (GST) (Salvio et al. 2016).Inhibition of acetylcholinesterase (AChE) activity in earthworms by GLY has been reported (Santos et al. 2011).However, the contribution of GLY toxicity in soil exposure assessments under eld conditions still needs to be improved.
The current study aimed to quantify the response of biomarkers in E. fetida using two experimental trials: in trial 1 (T1), the worms were exposed to high concentrations of GLY formulation under laboratory conditions, and in eld conditions (T2), where the worms were kept in soil collected from the agricultural eld from the agricultural regions known as Chenes in the state of Campeche, Mexico where a great variety of crops and a growing model of traditional Mayan and Mennonite agricultural production units, in addition to sorghum and soybeans, are grown.In both trials, biomarker involvement in biotransformation, oxidative damage, antioxidant enzymes, neurotoxicity, and GLY residues in soil collected from crops were evaluated.

Site description and soils of crops sampling
The locality named "Ejido Ich-ek" is located about 16 km from the municipality of Hopelchén (89°58 ′ 0 ″ W/19°43 ′ 60 ″ N) and has agriculture and beekeeping as its main activities covering an area of 49,680.00ha.After harvesting the crops, the pro le surfaces of crop soils (0-20 cm) were sampled using a metal shovel using a GPS for positioning.At each sampling site, approximately one kg of mixed soil samples from different crops was collected and labeled (Fig. 1).Soil samples were sieved (6.3 mm mesh) to remove larger soil aggregates and kept in well-labeled sampling bags and transported to the Ecotoxicology Laboratory of the Institute of Ecology, Fisheries and Oceanography in the Gulf of Mexico (EPOMEX), Campeche, Mexico.In the laboratory, the soils were dried, sieved (≤ 2 mm), and stored at 4° C until chemical and biological analyses were performed.The main physicochemical characteristics of the soils were clay 38%, silt 33% and 30%, pH 7.5 ± 0.1, organic matter 3.5%, and electrical conductivity 0.25 dS m − 1.

Glyphosate (GLY) analysis
Due to the solid binding capacity of GLY to the soil matrix, this herbicide is challenging to extract and achieve the best e ciency solely by extraction under alkaline conditions (NaOH).Due to the former fact, GLY in soil samples was extracted according to the procedure described by the ABRAXIS Glyphosate kit.Brie y, 10 g of homogenized dry soil was shaken with 25 ml of 1M NaOH for 30 min and centrifuged at 2500 g for 20 min.The supernatant was ltered through a 0.22 µm Durapore membrane lter, diluted (1:100) with Assay Sample Diluent, and extracts were stored in glass vials with Te on stoppers.The commercially available ELISA method by Abraxis LLC (Warminster, PA, USA; http://www.abraxiskits.com)was used for the immunoanalytical detection of GLY in soil.Measurements were carried out according to Rendón-von Osten and Dzul-Caamal (2017).The analytical detection was expressed as 1 ppm = 1 mg kg-1.

Earthworms and control (natural) soil
Eisenia fetida used in the study was cultured in the Ecotoxicology Laboratory of the EPOMEX Institute of the University of Campeche following the methods adapted in a non-contaminated natural soil (Jänsch et al. 2006;Kuperman et al. 2006;Colacevich et al. 2011; García-Santos and Keller-Forrer 2011).The specimens were collected from an area with no history of productive agro-activities in the Yucatan Peninsula.Natural and non-polluted soil was used for laboratory experiments because the physical and chemical properties of arti cial soils often do not represent the properties of natural soils.In addition, the natural and non-polluted soils from the same geographical region are representative of the cropping and polluted systems; in this way, the experiments under laboratory experiments have greater realism and make the data more useful.Other authors have also proposed using soils with tropical physicochemical characteristics (Bandeira et al. 2020).The soil was pedologically characterized to help with the data interpretation taking into account the possible effects of abiotic factors on enzymatic activities (clayey sand with 28.7%, sand 30.1%, 41.1% clay, organic matter: 3.7%, pH: 6.5, the electrical conductivity of 0.28 dS m-1).On the other hand, the worms were previously acclimated to the same temperature as the test (25 ± 2°C).The soil was adjusted to 30-60% of moistness; cow dung (5‰) was supplied as food, with a natural light-dark cycle, and to keep the moisture in worms' skin, the initial humidity of the soil was maintained by adding distilled water twice a week.Before the trial, healthy earthworms with adult clitella (0.275 g ± 0.0167 fresh weights, mean ± SD) were selected and placed on damp lter paper for 24 h to remove gut contents.

Exposure schemes
Trial 1 (T1): Acute toxicological test with glyphosate formulation under laboratory conditions Trial (T1) was performed using natural soil (Jänsch et al. 2006; Kuperman et al. 2006;Colacevich et al. 2011; García-Santos and Keller-Forer 2011; Bandeira et al. 2020) in 1,000 ml glass container with 500 g of dry soil at a controlled temperature (25 ± 2), humidity (60%) and a natural cycle of light and darkness.The GLY herbicide was added in its commercial presentation COLOSO TOTAL 360® Syngenta (360 g a.i / L-1 soluble in water) diluted in distilled water as follows: 1, 10, 100, and 500 mg kg-1.These concentrations were chosen based on the recommended eld application rate of 3-6 L a.i ha-1 (equivalent to 1440 g GLY a.i ha-1) and following on previous (Correia and Moreira 2010; García-Torres et al. 2014).Three replicates with ten clitellum earthworms of similar size and weight (0.275 g ± 0.0167 fresh weight) per container were exposed to each concentration, including the distilled water for the control group (OECD 1984).Acute tests were performed for 7 and 14 days without food and no addition of water to the soils because the necessary was provided at the beginning of the experiments.Survival was assessed at the end of the exposure period.Before biochemical analyses, earthworms were kept on moist lter paper for 24 h to allow bowel emptying.
Trial 2 (T2): Exposure to crops polluted soils samples A test was conducted according to standardized protocols (OECD 1984;Colacevich et al. 2011;Bandeira et al. 2020;Dzul-Caamal et al. 2020).Soils from the different polluted sites were adjusted with distilled water to a moisture content of 60% and pH (6.5 ± 0.5).For each soil replicate, ten worms with clitellum of similar size and weight (0.275 g ± 0.0167 fresh weight) were kept in 500 g of soil and placed in 1000 mL glass test containers with a perforated cap to prevent moisture loss.Two replicates per soil sample were used, including control of deionized water.The test containers were maintained in a climate-controlled room at 25 ± 2°C and a natural light-dark cycle.Distilled water was added twice weekly to keep moisture in the exposure media.Survival was assessed at the end of the exposure.Before biochemical analysis, the earthworms were kept on moist lter paper for 24 h as described previously.
An aliquot of the homogenate was transferred to a microtube to evaluate LPO levels.The remaining homogenates were centrifuged (12,000×g/20 min/ 4°C) in a microcentrifuge (Hermle Labnet Z216 MK) to obtain the S9 fraction (post-mitochondrial fraction) for determination of enzymatic activity (AChE, GST SOD, CAT, and GPx).Both fractions were stored at − 70° C until the biochemical assays were done.
Total and diluted protein concentration was measured using a Bio-Rad protein assay (BioRad, Richmond, CA) based on the protein-dye binding protocol of Bradford (1976), adapted to a microplate, using bovine γ-globulin as standard.
The decomposition product of lipid peroxidation (LPO) was determined by the formation of thiobarbituric acid reactive substances (TBARS) according to the method of Buege and Aust (1978) with microplate modi cations.150 mM Tris-HCl pH 7.4 buffer solution was added to 250 µL of the sample (without centrifugation).The sample was incubated at 37°C for 15 min.A TCA-TBA solution (0.375% thiobarbituric acid dissolved in 15% trichloroacetic acid) was added, then a thermal shock was induced in boiling water for 45 min.At the end of this time, it was cooled and centrifuged at 3,000 rpm for 10 min, and the absorbance at 535 nm was determined.LPO results were expressed as nmol TBARS/mg total protein, using the molar extinction coe cient (MEC) of 1.56x10 5 M/cm.Acetylcholinesterase (AChE, EC 3.1.1.7)activity was measured using acetylthiocholine (ATC) as substrate according to the colorimetric method of Ellman et al. (1961) with some microplate modi cations.Each reaction mix included 50 µl enzyme extract diluted in 250 µl reaction solution (0.25 mM ATC, 0.4 mM DTNB and 0.1 M pH 7.5 phosphate buffer).The enzyme activity was determined kinetically at 414 nm.AChE activities were expressed as nmol ATC hydrolyzed per minute per mg of protein using the MEC (13.6 mM cm − 1 ).
Glutathione S-transferase (GST, EC 2.5.1.18)activity was evaluated by the method of Habig et al. (1974), adapted to a microplate using the conjugation of GSH with 1-chloro-2,4-dinitrobenzene (CDNB).Absorbance was recorded at 340 nm (25 •C) for 3 min and expressed as nmol CDNB/min/mg protein (ε = 9.6 mM cm − 1 ).Superoxide dismutase (SOD, EC 1.15.1.1)activity was determined by the method described by Misra and Fridovich (1972), which is based on inhibiting autooxidation of adrenaline to adrenochrome at alkaline pH.30 µL of supernatant in the microplate was added to 220 µL of reaction solution (0.3 mM adrenaline in carbonate buffer [50 mM sodium carbonate and 0.1 mM EDTA], pH 10.2).A Blank of water was used as a reference.Absorbance was read at 480 nm after 20 s and 2 min.The activity was determined using the MEC of SOD (21 M cm-1).Results were expressed as nmol SOD/min/mg protein.
Catalase (CAT, EC 1.11.1.6)activity was determined by the method described by Radi et al. (1991), estimated by the dismutation of hydrogen peroxide (H 2 O 2 ) at 240 nm after 0 and 60 s.To 200 µL of supernatant was added to a microplate, and 100 µL of reaction solution of 20mM of H 2 O 2 contained in isolation buffer (0.3M sucrose, 1mM EDTA, 5mM HEPES, and 5mM KH 2 PO 4 ).CAT activity was calculated as mmol of H 2 O 2 consumed per minute per mg of protein, using the MEC of 0.043 mM cm − 1 .
All these spectrophotometric methods were performed using a microplate reader (Multiskan Spectrum, Thermo Scienti c).

Data analysis
For the biomarker data, at least four replicates were analyzed for each earthworm sample collected from the soil.Each consisted of 10 earthworms incubated in 500 g of soil, and all values were represented as mean ± standard deviation (SD).Data were tested for normality (Kolmogorov -Smirnov's test).For the acute toxicological test with GLY formulation under laboratory conditions, means were compared by oneway ANOVA (parametric) followed by the Dunnett multi-comparisons test were performed to discriminate signi cant changes in the biochemical parameters following exposure (1, 10, 100, and 500 mg a.i kg-1) compared to the corresponding control.For the exposure to soils collected from crops (S1 to S9), the mean comparisons were conducted using ANOVA followed by Tukey's post hoc test.Alternatively, Kruskal-Wallis, followed by Dunn's test, was utilized for non-parametric data.For all statistical purposes, the criterion of signi cance was set at p < 0.05.GraphPad InStat software (GraphPad Software, San Diego, USA) was used for the analysis.
In order to integrate all biomarker data in both experiments (GST, SOD, CAT, GPx, LPO, and AChE) into a general stress index, a method for calculating the Integrated Biomarker Response (IBRv2) based on reference deviations.We proposed a slight addition to IBRv2, named the general IBRv2, which is calculated as the sum of all IBRv2s involved in a particular situation (gIBRv2 = IBRv2 i = 1 ai = n) to integrate the total effects as a single value (Dzul-Caamal et al. 2016 a,b).The control group was considered as a reference.Biomarker deviation indices for each exposure condition were reported in star plots.The areas above and below 0 indicate biomarker induction and inhibition, respectively.
To determine the most relevant variables that explain the variance between the analyzed soils and the possible relationship among the biological biomarkers, a principal component analysis (PCA), followed by the analysis of Agglomerative Hierarchical Clustering (AHC), was conducted using the software XLSTAT-Ecology v.2018.5 (Addinsoft, 2018).AHC was used to identify class differences and cluster the samples with similar biomarkers and GLY content.Results are presented in a dendrogram where steps in the hierarchical clustering solution and values of distances between clusters (squared Euclidean distance) are represented.

Results
Trial 1 (T1).Acute toxicological test with glyphosate formulation using unpolluted soils No mortality was observed, and all the earthworms remained in the containers during exposure to the GLY formulations.However, biochemical measurements revealed signi cant differences between treatments (repeated measure ANOVA, P < 0.05), whatever the length of exposure (7 or 14 d).
GST activity showed a signi cant increase dependent on test concentrations at 7 and 14 d of exposure to the glyphosate formulation compared to the control (Fig. 2A).In both exposure times, the higher activity was found at 100 and 500 mg a.i.kg-1 (Fig. 2A).
Exposure to the GLY formulation also increased the activity of antioxidant enzymes for 7 and 14 d (Fig. 2).The peak of SOD activity occurred at concentrations of 10 and 100 mg a.i.kg-1 at 7 d and concentrations of 1 and 10 mg kg-1 at 14 d of exposure compared to control treatments (Fig. 2B).However, after 7 and 14 d of exposure, there was a decrease in its activity, mainly in the concentration of 500 mg a.i.kg-1 (Fig. 2B.).In the same way, a similar behavior could be observed in the activity of the enzymes CAT and GPx which reached their higher activity (P < 0.001) at concentrations of 100 mg a.i.kg-1 at 7 and 14 d of exposure (Fig. 2C and D).However, in the case of CAT, a general decrease in activity was observed at 14 d.In contrast to the activity of GPx, which increased its activities in all concentrations at 14 d of exposure, the CAT decrease indicates the effect by the concentrations of the GLY formulation (Fig. 2C and D).LPO contents measured as levels of thiobarbituric acid reactive substances (TBARS) in earthworms are shown in Fig. 2E.The commercial formulation of GLY induced an increase in TBARS levels (P < 0.001) during the exposure time.However, the maximum increase in TBARS was observed at 500 mg a.i.kg-1 at 7 and 14 d, respectively (Fig. 2E).This increase remained constant.It even increased in all GLY concentrations at 14 d of exposure, indicating the initiation and development of oxidative stress and tissue damage elicited by GLY formulations in earthworms.
Regarding AChE, the activity of this enzyme was signi cantly (P < 0.05) inhibited from earthworms, as shown in Fig. 2F.At the 7 d, the higher inhibition (28%, and 27%, respectively) occurred at the lowest concentrations of 1 and 10 mg a.i.kg-1.However, after 14 d of exposure, these inhibitions increased to 31%, 40%, and 37%, respectively, at concentrations of 1, 10, and 100 mg a.i.kg-1, concerning the control group (Fig. 2F).After 7 and 14 d, in all concentrations tested, AChE activity showed a "recovery effect" increasing this response at 500 mg a.i.kg-1 (Fig. 2F).

General Integrated biomarker response (gIBRv2)
The highest value of gIBRv2 (29.80) for the responses of biomarkers evaluated in E. fetida was found in earthworms exposed at 14 d to soils with concentrations of 1, 10, and 100 formulated GLY (mg a.i kg-1) (Fig. 3A).In general, the "A" values obtained by the IBRv2 method for biomarkers (SOD, CAT, GPx, and TBARS) were higher (7.45) than those observed in the control group.While the GST and AChE biomarkers presented negative values of IBRv2 (-0.964 and − 2.941) in the test concentrations of 1 to 1000 formulated GLY (mg a.i kg-1) (Fig. 3B).

Glyphosate concentration in crops soils
The GLY concentrations determined in the nine sampled crop soils are shown in Fig. 4. The GLY content ranged between 0.50 and 0.92 mg kg-1.In particular, sites S7, S8, and S9 presented higher concentrations of GLY with 0.92, 0.86, and 0.87 mg kg-1, respectively.Only S1 (Soil without crop) had lower values of 0.51 mg kg-1 of GLY.Based on the values of GLY in the different soils, the accumulation level of this herbicide had a decreasing trend in the following way in the crops: soybean > maize > papaya.

Biomarker responses
No earthworm mortality was recorded during the exposure of the soils collected in the eld (T2).However, acute exposure to soils appeared to have induced oxidative stress, signi cantly increasing antioxidant enzyme activity and augmented lipid peroxidation.After 14 days, there were signi cant differences (P < 0.05) in the GST activities of earthworms exposed to eld-collected soils compared to the control (Fig. 5A).The S7 and S8 (Soybean and Maize) exhibited the highest GST activity (160 and 146%, respectively) (Fig. 5A).
The speci c activity of antioxidant enzymes in earthworms exposed to agricultural soils is shown in Fig. 5. SOD, CAT, and GPX activities were relatively sensitive in the nine soils collected after 14 d compared to controls.However, only in the S7 (Soybean), the activity of the antioxidant enzymes increased signi cantly (P < 0.0001) compared to the control, reaching activities of 177%, 228%, and 208% in the activities of SOD, CAT, and GPx (Fig. 4B, C and D).While in soils S3, S5, S8, and S9 (Maize, Soil preparing for crops, Maize and Papaya), the signi cant responses (P < 0.05) of these enzymes kept their activities constant (164%, 173%, and 159%, respectively) (Fig. 5B, C and D).
LPO content in earthworms was also affected by exposure to eld-collected soils.Figure 5E shows an increasing trend of lipid peroxidation levels in all agricultural soils up to 14 d.However, only soil 7, 8, and 9 exhibited the highest levels (169%, 155%, and 142%) and signi cant (P < 0.0001) levels of lipid peroxidation.In comparison, the lowest values of lipid peroxidation levels (120%) were observed in S2 and S3 (Chihuahua squash, Habanero pepper) (Fig. 5E).

Integrated biomarker response and multivariate analysis
The effects of combined glyphosate on the biomarkers responses of earthworms were analyzed by General Integrated biomarker response (gIBRv2), Multivariate analysis (PCA), and Agglomerative Hierarchical Clustering (AHC).The highest value of gIBRv2 (82.14) for biomarker responses evaluated in E. fetida was found in earthworms exposed at 14 d (Fig. 6A).In general, the "A" values obtained by the IBRv2 method for biomarkers (TBARS, SOD, CAT, and GPx) were higher than those observed in the control group (13.12, 10.60 and 9.85) and were present in S7, S8, and S9 (Soybean, Maize, and Papaya) (Fig. 6B).However, GST and AChE presented negative values (-0.88 and − 3.19) (Fig. 6A).
In general, PCA showed two main factors that explain 93.89% of the data set's variability, indicating discrimination between agricultural soils.The two components represented 89.19% (PC-1), and 4.70% (PC-2) of the variability, respectively, and were represented in the Euclidean biplot (Fig. 7A).In the Euclidean biplot, the exposed and control groups clustered in different quadrants.The distribution is higher in the rst axis, where the main contributions were presented in soil designed for maize crops (S8), which positively grouped all the biomarkers (GST, TBARS, and SOD) and were in uenced by GLY.While the soils designed for soybean (S7), papaya (S9), and Soil preparing for crops (S5), grouped the activities of CAT and GPx, but were not in uenced by GLY.While AChE activity was clustered in Habanero pepper crop soils (S4 and S6), but also not in uenced by GLY (Fig. 7A).
Agglomerative Hierarchical Clustering (AHC) for grouping sample locations (Fig. 7B) was used to overview the similarities between soil GLY concentrations and biomarker response in E. fetida.HCA is a technique often used with PCA to study the structure of a data set.The dendrogram of the cluster analysis to the scores of the rst two principal components is presented in Fig. 7B.From this analysis, it is evident that crop soils to which E. fetida were exposed gathered according to biomarker response coe cients (GST, SOD, CAT, GPx, LPO, and AChE).The dendrogram is the result of the cluster analysis and represents three groups.The sample soils are named according to the type of crop they represent.C1 (S1, S2 and S4); C2 (S3, S5, S6, S8 and S9) and C3 (S7).However, only group C2 (S5, S8, and S9), which groups different crop soils, is characterized by high variance related to high GLY concentrations, con rming the conclusion previously obtained by PCA (Fig. 7B).

Trial 1 (T1). Acute toxicological test with glyphosate formulation Biomarker responses
The results of the trial (T1) performed with no polluted soils showed no evidence of mortality in E. fetida in soils treated with glyphosate formulation, even at the highest concentration (500 mg kg-1), which is similar to that reported in another study in worms that tolerated high concentrations of 1,000 to 5,000 mg ).However, these authors reported a considerable decrease in mean body weight, serious toxic effects on reproduction, signi cant anatomical changes in exposed earthworms, increases in enzyme activities, and decreases in microbial populations.Therefore, using glyphosate may not directly harm them, but it can cause severe effects and even long-term mortality in earthworms (Piola et al. 2013;Santadino et al. 2014).
Earthworms are frequently exposed to pesticides and therefore depend on e cient detoxi cation systems for their survival (Givaudan et al. 2014).The activities of detoxi cation enzymes such as glutathione Stransferase (GST) have been used as biomarkers of phase II detoxi cation, antioxidant defense, and amelioration of the oxidative stress effects of exposure to organic and inorganic compounds in earthworms (Aly and Schröder 2008;Givaudan et al. 2014).Our results showed a signi cant increase in GST activity in E. fetida dependent on the test concentrations of the glyphosate formulation (1-500 mg a.i.kg-1), which remained constant up to 14 d of exposure.The increased GST activity in E. fetida tissues could be a protective measure against glyphosate toxicity.According to Aly and Schröder (2008), the role of GST is to detoxify electrophilic intermediates through conjugation with glutathione, as also found by Contardo-Jara et al. ( 2009), which documented increases in GST activity in Lumbriculus variegatus caused by Roundup glyphosate formulations.Givaudan et al. (2014) reported that acute exposure of Aporrectodea caliginosa to 2.5 mg g-1 of active glyphosate leads to accelerated activation of the detoxi cation enzyme GST.Pochron et al. (2021) documented that E. fetida responds to prolonged or short exposures by generating a response in the survival time of the worm stress test.Current results and previous reports suggest that the worms nd the rst seven d of exposure the most physiologically challenging.
In our study, the activity of antioxidant enzymes in E. fetida showed signi cant increases depending on the concentrations of the glyphosate formulations and the exposure time (7 and 14 d).These responses may be due to the generation of ROS (Halliwell and Gutteridge 2015).It is documented that worms exposed to glyphosate formulations can suffer alterations in some components of the antioxidant defense system.Contardo-Jara et al. ( 2009) found that 96-h exposure to glyphosate and its formulations at concentrations ranging from 0.05 to 5 mg L-1 elicit increased SOD, CAT, and GST activity in the annelid Lumbriculus variegatus.Marcano et al. (2017) reported signi cant effects for all biochemical markers evaluated in Eisenia sp, treated with glyphosate formulation.Antioxidant defenses (GSH, GPx, and GST) generally increased during the 7 and 21 d exposure.These responses suggest that the active ingredient and/or its chemical additives of glyphosate exert their toxic effects on Eisenia sp. by altering the cellular antioxidant defenses, inducing a condition of oxidative stress.
On the other hand, our results also showed that the highest activities of these antioxidant enzymes occurred at the lowest concentrations of glyphosate (1 to 100 mg a.i.kg-1).The higher concentration of glyphosate (500 mg a.i.kg-1) produced a slight decrease in SOD, CAT, and GPx activities compared to the control at 7 and 14 days.This process could be explained by the body's ability to partially or fully overcome the stress resulting from exposure.However, the presence of high levels of ROS, particularly the overproduction of O 2 •, becomes an inhibitor of CAT and GPx (Lushchak et al. 2009).
One consequence of glyphosate metabolism accompanied by insu cient antioxidant pathways is lipid peroxidation (LPO), one of the principal indicators of oxidative stress that can be generated after exposure to pollutants such as herbicides (Marcano et al. 2017).Therefore, in this study, measurements of LPO levels were necessary to verify the effects of glyphosate on exposed E. fetida.Our results showed consistent signi cant increases in induced LPO levels at 7 and 14 d of exposure to the glyphosate formulation.These ndings were similar to those documented by Zhou et al. (2013), who documented increased LPO levels in E. fetida exposed for 28 d to glyphosate and a mixture of glyphosate with copper.Marcano et al. (2017) stated signi cant effects of LPO on Eisenia sp, treated with glyphosate formulation for 7 and 21 d of exposure.In our study, one explanation for the increase in LPO may be that glyphosate depletes the GSH content of the tissues, which may increase the risk of oxidative stress and decrease antioxidant enzymes such as SOD, CAT, and GPx, the rst barriers against ROS (Halliwell and Gutteridge 2015).
In addition to broad-spectrum biomarkers, such as those involved in response to oxidative stress, assessing other biomarkers, such as AChE activity, is a priority.AChE is an enzyme that has crucial functions in the nervous system and has been used to diagnose the toxicity of organophosphate (OP) and carbamate (CM) pesticides in animals and humans (Mwila et al. 2013) and in sh species (Menéndez-Helman et al. 2012; Sandrini et al. 2013).Although little studied, the inhibition of AChE by GLY was demonstrated in earthworms (Santos et al. 2011;Zhou et al. 2013).In the current study, AChE activity in E. fetida statistically decreased (40%) with increasing concentrations of glyphosate in the soil.These data con rm that AChE could be considered a target enzyme of glyphosate on the nervous system of this species and coincide with the results reported by other authors in E. fetida and other earthworm species (Santos et al. 2011;Zhou et al. 2013).Tarouco et al. (2017) observed the inhibition of two cholinesterase isoforms (acetylcholinesterase and propionylcholinesterase) in the estuarine polychaete Laeonereis acuta exposed to both concentrations of Roundup after 96 h.However, the inhibition of AChE in our study only happened at low concentrations of glyphosate (1, 10, and 100 mg a.i.kg-1) at 7 and 14 d of exposure.However, at the higher concentration of glyphosate (500 mg a.i.kg-1; Fig. 2), an increase in the recovery of this enzyme was observed.The recovery in AChE could mean that the earthworm "turn-on" its metabolic battery through the gene transcription for new AChE synthesis or by other detoxifying mechanisms for reducing the internal dose of the pesticide, such as the activation of other cholinesterases (Reiner 1971).Another via involved in AChE recovery is simply due to adaptation (hormesis).The hormesis response to AChE activity has been documented in earthworms E. fetida and E. andrei exposed to temephos and glyphosate (Hackenberger et al. 2008;Pochron et al. 2021).
The integrated biomarker response (IBRv2) approach has been successfully applied to study biological responses in several eld and laboratory studies (Dzul-Caamal et al. 2016 a,b).In the present study, gIBRv2 discriminated exposure conditions based on biomarker response.The data integration updated the overall picture obtained by evaluating individual biomarker responses and the severity of exposure conditions, standardized values in star plot areas.The results of the current study show that the gIBRv2 index coincided with the individual response of the biomarkers evaluated.High concentrations of GLY (100 and 500 mg a.i kg-1) coincided with the higher inductions of SOD, CAT, and GPx and a signi cant increase in LPO content, indicating the association indicating the turn-on oxidative stress response.This response suggets that earthworms have the ability to withstand the challenge of oxidative stress by activating antioxidant defenses.There are no studies was evaluated IBRv2 in earthworms exposed to controlled conditions to glyphosate; however, Zhu et al. (2020) showed induction of reactive oxygen species, antioxidant enzyme activities, detoxifying enzyme activity, malondialdehyde content, and 8hydroxydeoxyguanosine content in E. fetida exposed to chlorpyrifos.Subchronic toxicity was quanti ed using the integrated biomarker response (IBR), which highlighted that the toxicity of chlorpyrifos in arti cial and natural soils was not the same.The obtained results from the GLY analysis in the sampled crop soils showed the presence of this herbicide in nine of the monitored soils with a variation of 0.50 to 0.92 mg kg-1.The higher concentration of GLY was found in S7 (Soybean), S8 (Maize), and S9 (Papaya), respectively.Currently, in Mexico, there is no maximum level of glyphosate residues in soils; however, all glyphosate residue levels found in the present study were above the maximum residue level (MRL) for soils according to the European Union Pesticide Database for GLY of 0.2 mg kg-1 and 0.5 mg kg-1 (Silva et al. 2018).Based on the content of glyphosate in the different crops, the degree of pollution of this herbicide was found as follows from higher to lower Soybean > Corn > Papaya.These values indicate that glyphosate residues are present in both crop plants genetically modi ed to be tolerant to GLY, such as soybeans and non-resistant plants, as were the case of corn and papaya exceeding the MRL values of 0.4-8.8mg kg-1 for GLY (Cuhra 2015 In Ich-Ek locality has been documented concentrations of glyphosate in groundwater (1.42 µg L-1).These concentrations indicate an unusually high eld application rate of glyphosate and the risks in other environmental compartments due to surface erosion and leaching (Rendón-von Osten and Dzul-Caamal 2017).

Biomarker responses
As observed in T2 of the current study, the assessment of GST activity is a suitable tool for evaluating agricultural soil contamination.The exposure of E. fetida to S5 (Soil preparing for crops), S7 (Soybean), S8 (Maize), and S9 (Papaya), which presented higher levels of GLY (0.73, 0.92, 0.86 and 0.87 mg kg − 1), increased the activity of Glutathione S-transferase concerning the control group.The increase in GST activities in earthworm tissue could be a protective measure against the toxicity of pesticides such as glyphosate.These results agreed with those documented by Owagboriaye et al. (2020), which found higher GST and LDH activities in three species of earthworms (Alma millsoni, Eudrilus eugeniae, and Lilyodrilus violaceus) exposed to GLY compared to those not exposed.Marcano et al. (2017) documented increases in the GST in Eisenia sp.exposed to acute and chronic concentrations of a commercial formulation of GLY.
Many pesticides, such as GLY, have been documented to be potent inducers of ROS in earthworms (Bailey et al. 2018).However, to prevent oxidative damage (ROS), cells are protected by antioxidant enzymes, including SOD, CAT, and GPx, acting together to reduce the effect caused by extra active oxygen (Halliwell and Gutteridge 2015).In this study, we observed that exposure to S3 (Maize), S5 (Soil preparing for crops), S7 (Soya), S8 (Maize), and S9 (Papaya) induced signi cant increases in SOD, CAT, and GPx activity.These results are similar to those Marcano et al. (2017) reported, who documented increases in antioxidant defenses (GPx, GST, and reduced glutathione levels) in Eisenia sp.exposed to commercial formulation of GLY.The induction of these enzymes may re ect an adaptation of earthworms for ROS removal (O 2 • and H 2 O 2 ) (Shi et al. 2016).It has also been documented that the response of SOD, CAT, and GPx varies according to ROS generation, and organisms could respond to oxidative stress differently (Sandrini et al. 2013).This response was evidenced in our study, where the activity of GPx was lower than CAT may be because CAT is the main enzyme that reacts in the oxidative state during the rst days of exposure to inorganic compounds as a compensation mechanism (Xiong et al. 2013).
The obtained results show that the contents of LPO increased in the worms exposed to S6, S7, S8, and S9, indicating that the ROS generated by these contaminants are not entirely detoxi ed by antioxidant enzymes (Halliwell and Gutteridge, 2015).As a consequence, oxidative stress is generated by exposure to the GLY residues present in these crop soils.Consequently, the presence of LPO in the worm E. fetida is a potential biomarker of oxidative stress due to environmental pesticide exposure.
Our study found an AChE inhibition of 61% in E. fetida after 14 d of exposure to agricultural soils, with no mortality.This survival sensitivity to AChE inhibition has also been recently documented in other earthworm species (Allolobophora chlorotica and Aporrectodea caliginosa).It suggests that the toxic action of pesticides may involve molecular or cellular sites of action beyond AChE (Rault et al. 2007).

In uence of glyphosate concentrations in soils on biomarker responses
In monitoring studies, the use of integrated biomarker response (IBR), Principal Component Analysis (PCA), and Agglomerative Hierarchical Clustering In the current study, an excellent visual relationship between the glyphosate gradient in crop soils observed in the soils and the IBR variation in the different sampled soils was detected.Similarly, data integration using PCA analysis revealed a clear association between the intensity of biomarker responses and GLY concentrations, discriminating the level of contamination of each soil.Was found that most biomarker responses were positively grouped with GLY levels, con rming the higher toxicity elicited by soils S7, S8, and S9 (soybean, corn, and papaya).This nding indicates oxidative stress in the earthworms early at 14 d of exposure.Similar results have been documented for others pollutants for ROS levels, oxidative and DNA damage, increases in enzyme activities, and changes in body weight in earthworms (Liu et  The GLY metabolism can explain current ndings by earthworms; during this process, higher levels of free radicals induce oxidative stress that could collapse the activity of antioxidant defenses (Cattani et al. 2017).In addition, the inhibition of AChE induced by GLY in E. fetida can be attributed to cysteine residues near the enzyme's active site (Frasco et al. 2007), compromising in this way, the catalytic activity or denaturation of the enzyme.However, it should be noted that the interpretation of eld results is always a very complex operation since many factors can in uence the analyzed variables in an

Conclusions
The results of this study con rm that GLY cause adverse biological effects such as oxidative stress and increase the activities of GST, SOD, and levels of LPO in Eisenia fetida, both under controlled conditions as in GLY-polluted soils.This response was more sensitive in E. fetida exposed to agricultural soils polluted with GYL than in GLY-enriched unpolluted soils.However, AChE inhibition was more noticeable in exposures to GLY contaminated soils and presented hormesis, this could indicate may be related to secondary or indirect effects of GLY on earthworm physiology and not through a speci c action mechanism in contrast with toxic effects elicited by organophosphates or carbamates pesticides.
Additionally, the soil analysis revealed the presence of GLY, particularly in soybean cultivation in the Yucatan Peninsula, as in other sites worldwide.Finally, the integrated evaluation indicated that laboratory and eld studies describe a good relationship between biomarkers and polluted levels for different contaminated sites, which are needed to prevent, control, reduce environmental contamination and minimize adverse effects on natural populations in Yucatan Peninsula.

Trial 2 (
T2). Exposure to agricultural soils samples Glyphosate residue levels in cropping soils collected In the municipality of Hopelchén, Campeche, the agro-industrial model has advanced more than in any other agricultural area of the Yucatan Peninsula.Pesticides are used intensively in corn, sorghum, and tomato crops, among others, and the use of GLY has been emphasized with the implementation of transgenic crop species such as soybeans (Villanueva-Gutiérrez et al. 2014; Rendón-von Osten and Dzul-Caamal 2017).
; Dill 2005; Villanueva-Gutierrez et al. 2014).These results are similar to those Aparicio et al. (2013) reported, who documented concentrations of 0.35 and 1.5 mg kg-1 of GLY in Argentina's soybean, corn, and cotton soils.Other studies have documented GLY residues detected in soils from crop production elds in the United States ranging from 0.025 to 1.0 mg kg-1 (Laitinen et al. 2006; Benbrook et al. 2016).Peruzzo et al. (2008) reported concentrations of 0.5 and 5.0 mg kg-1 of glyphosate in soils associated with the direct sowing of soybean cultivation in the northern Pampas Region of Argentina.In Europe, the upper levels of GLY in soil vary considerably between countries.For example, up to 2.05 mg kg-1 of glyphosate has been found in vineyards in Portugal (Silva et al. 2018).Syan et al. (2014) found concentrations ranging from 0.057 to 0.011 mg kg-1 of glyphosate in soil used for transgenic canola production in Quebec, Canada.
uncontrollable way (Dzul-Caamal et al. 2016, 2020).Pochron et al. (2021) documented that earthworms live with and withstand multiple doses of GLY-based herbicides at the eld level.

Figure 3 General
Figure 3

Figure 6 A
Figure 6