Agrochemical Mixtures and Amphibians: The Combined Effects of Pesticides and Fertilizer on Stress, Acetylcholinesterase Activity, and Bioaccumulation in a Terrestrial Environment

Abstract

Tank mixtures are popular within the agricultural community because they are time- and cost-effective, but field applications leave nontarget organisms at risk of exposure. We explored the effects of a common herbicide (atrazine and alachlor) and fertilizer (urea) tank mixture on juvenile frog corticosterone stress levels, acetylcholinesterase (AChE) activity, and pesticide bioaccumulation. Single agrochemical or tank mixtures were applied to terrestrial microcosms, and then individual Southern leopard frog (Lithobates sphenocephala) juveniles were added to microcosms for an 8-h exposure. Afterward, frogs were transferred to aquatic microcosms for 1 h to monitor corticosterone prior to euthanasia, brain tissues were excised to evaluate AChE, and tissue homogenates were analyzed for pesticide bioconcentation with gas chromatography–mass spectrometry. Atrazine significantly increased corticosterone in frogs, particularly when combined with alachlor and urea. Atrazine increased AChE and urea decreased AChE, although no interactive effects of chemical combinations were discernible. Relative to their individual treatments, the complete tank mixture with all 3 agrochemicals resulted in 64%greater bioconcentration ofatrazine and 54%greater bioconcentration ofalachlor in frog tissues. Our results suggest that agrochemical mixtures as well as their active ingredients can lead to altered stress levels and impaired physiological responses in amphibians. An improved understanding of the effects of co-exposure to environmental contaminants in amphibians is important in assessing the ecological risks these compounds pose.

INTRODUCTION

Agrochemicals, primarily pesticides and fertilizers, are used regularly in an effort to control damaging crop organisms and increase crop productivity and yield. In the United States alone, estimated pesticide use on agricultural lands increased from 97522 metric tons in the 1960s to nearly 250000 metric tons as of 2010 (Osteen and Fernandez-Cornejo 2013). With the advent of pesticide-resistant crops and a growing population, their use is projected to continue to increase globally. Worldwide nitrogen and phosphorus fertilizer use estimates, excluding the former Soviet Union, totaled over 100 million metric tons by the mid-1990s, 7 and 3.5 times greater, respectively, relative to the 1960s (Tilman et al. 2002). The widespread use of chemicals in agriculturally intensive landscapes poses a significant risk for nontarget organisms, both during and after application. Amphibians are one nontarget species group that traverse long distances, to and from breeding ponds, often through croplands, leaving them vulnerable to agrochemical exposure (Mann et al. 2009; Berger et al. 2012; Fryday and Thompson 2012; Lenhardt et al. 2015; Swanson et al. 2018). Population loss among many amphibian taxa is a prevalent and well-known conservation concern (e.g., Stuart et al. 2004; Pounds et al. 2006; Alford et al. 2007). Understanding the potential role that agrochemicals play in amphibian declines is essential for both short- and long-term population sustainability.

Pesticide exposure alone is known to have harmful impacts on amphibians, both during the embryonic/larval aquatic stages and after metamorphosis. Many studies have investigated the negative effects of pesticide exposure on embryos and developing tadpoles, including deformities, arrested development, mortality, endocrine and immune disruption, and increased disease susceptibility, among others (e.g., Bridges 2000; Hayes et al. 2006; reviewed in Mann et al. 2009; Baker et al. 2013). In juvenile and adult amphibians, pesticides have been shown to inhibit cholinesterase production, alter feeding behavior, induce changes in weight, and ultimately result in death (reviewed in Brühl et al. 2011). The use of more than 40 different fungicide, herbicide, and insecticide products has been reported during a 1-yr period on winter crops and maize in Germany. Temporal overlaps between pesticide application and migration events by various amphibians were recorded, with as few as 1% to as many as 87% of migrating individuals/species coincident with at least 1 chemical application (Lenhardt et al. 2015).Therefore, variations in the timing of breeding migrations, along with crop- and pest-specific pesticide applications, leave many amphibian species more susceptible to pesticide exposure in the field relative to exposures in aquatic environments (Lenhardt et al. 2015; Swanson et al. 2018). The uncertainties around species-specific exposure patterns as well as a lack of knowledge of the mechanisms of pesticide toxicity for many registered compounds in these nontarget species present a challenge for amphibian risk assessment.

Although fertilizer effects on amphibians have been less well studied, these compounds also contribute to impaired larval development, mortality, decreased feeding activity, and disruption of physiological activity such as thyroid hormone production and endocrine disruption, among other adverse effects (reviewed in Rouse et al. 1999; Mann et al. 2009; Baker et al. 2013). For postmetamorphic amphibians, mortality, reduced feeding, and respiratory distress have been reported after fertilizer exposure (reviewed in Rouse et al. 1999; Mann et al. 2009). Choice-based laboratory studies using urea and ammonium nitrate fertilizers detail mixed results on avoidance behaviors among amphibians (e.g., Marco et al. 2001; Hatch et al. 2001; Egea-Serrano et al. 2011; Mann et al. 2009; Gaglione et al. 2011), which suggests that amphibians may not be able to detect and evade fertilizers in a field setting. Furthermore, fence-trapping efforts in Germany, throughout a large agricultural landscape, showed high amphibian movement across recently fertilized fields, with early spring–breeding amphibians being at greater risk of fertilizer exposure (Berger et al. 2012). Migrating amphibians may also be vulnerable to exposure to numerous fertilizers simultaneously. Across 3 farms studied to better understand fertilizer management plans, a total of 11 different compounds were used between 2007 and 2008, thereby increasing the number of fertilizers to which amphibians could be exposed (Berger et al. 2012). Croplands are important corridors for amphibian dispersal but present serious risks given the temporal overlap with fertilizer applications at critical stages of their life histories.

For decades, large-scale crop production has relied on the simultaneous application of pesticides and fertilizers through tank mixtures. Although mixing may be a preferred method of agrochemical application in many instances, simultaneous chemical exposure to target and nontarget species may result in synergistic or ameliorative effects. Boone et al. (2005) reported indirect ameliorative effects of co-exposure to nitrate and carbaryl in green frog tadpoles, in which increases in algal food resources through nitrate addition were negated by carbaryl. Similarly, Smith et al.(2011) reported a corrective effect of nitrate and malathion exposure on wood frog development: the presence of nitrate decreased the negative biological impacts of malathion. Furthermore, leopard frog tadpoles were not impacted by the combination of nitrate and atrazine, although nitrate alone decreased developmental rates (Allran and Karasov 2000). Studies examining co-exposure of juvenile and adult amphibians to pesticides and fertilizers are lacking, although warranted, given the individual effects across all life stages and the combined effects in tadpoles as detailed above.

The co-application and residence time of residual agrochemicals pose a threat to amphibians throughout agricultural areas. In an effort to close the knowledge gaps on the combined effects of pesticides and fertilizers on terrestrial phase amphibians, the present study was designed to explore active ingredient uptake, as well as potential biological indicators of agrochemical exposure, including corticosterone stress response and acetylcholinesterase (AChE) activity, in Southern leopard frogs (Lithobates sphenocephala). Postmetamorphic, juvenile frogs were exposed to 3 agrochemicals (atrazine, alachlor, and urea [2 herbicides and 1 fertilizer, respectively]) as individual or combined chemical treatments. We hypothesized a cumulative response in bioaccumulation, an increase in corticosterone, and a decrease in AChE activity among frogs when exposed to both herbicides simultaneously. With the addition of the urea fertilizer to the herbicides, a more pronounced response among frogs was predicted, whereby urea would facilitate the accumulation of herbicides, ultimately leading to higher corticosterone levels and a further decrease in AChE activity.

MATERIALS AND METHODS

Chemicals and soils

All solvents and urea (>99%purity) were obtained from Fisher Scientific. Atrazine and alachlor (>98% purity) were purchased from Chem Service.

Soils were collected from a grassland plantation at the Chester River Field Station of Washington College, Chestertown (MD,USA)in June 2016 and stored in a walk-in cold room at 4°C. Prior to experimental use, soils were passed through a 2-mm sieve to remove large pieces of debris. The soil was classified as a Unicorn–Sassafras loam and is typical of soils in agricultural fields with low organic matter.

Amphibian rearing and care

Southern leopard frogs (L. sphenocephala) were collected as egg masses from vernal pools in Queen Anne’s County (MD, USA) and transported to Washington College. Eggs were hatched indoors in 10-gal aquaria, and tadpoles were transferred to outdoor 110-gal polyethylene tanks filled with aged tap water after they had reached Gosner stage 25 (the free feeding stage; Gosner 1960). Tadpoles were fed Tetra Min® tropical fish flakes ad libitum through metamorphosis. At metamorphosis, all juveniles were transferred to outdoor 110-gal polyethylene tanks lined with moist sphagnum moss to simulate a terrestrial habitat and fed purchased crickets for 30 to 90 d until experimentation.

Agrochemical exposures

Juvenile leopard frogs were divided randomly across 8 chemical treatments: control (C), atrazine (Z), alachlor (L), urea (N), atrazine + alachlor (ZL), atrazineþurea (ZN), alachlorþurea (LN), and the full chemical combination of atrazine + alachlor + urea (ZLN). Eighteen frogs were randomly assigned to each chemical treatment for a total sample size of 144 across the study. After exposure, 12frogs/treatment (n=96) were assessed for corticosterone levels using a waterborne assay, followed by brain AChE activity and active ingredient body burden analysis using the methods detailed below (in the sections entitled Corticosterone analysis, AChE analysis, and Soil and amphibian agrochemical extraction). The remaining 6 frogs/treatment (n=48) were immediately euthanized to verify active ingredient tissue accumulation prior to depuration. Due to the time required to apply chemical treatments to soils as well as the time needed to transition from terrestrial to aquatic microcosms, the experiment was conducted over a 2-d period in September 2016.Allfrogs subject to corticosterone analysis were tested on the same day (n=96 frogs on day 1).

Agrochemical application and subsequent exposure followed the methods outlined in Van Meter et al. (2016). Briefly, experimental chambers were 0.94-L Pyrex® glass bowls lined with 150g soil. Pesticide active ingredients were applied at the labeled application rate: atrazine μg/cm². Urea was applied at 2.2 μg/cm=23.60μg/cm₂, which is within² and alachlor =34.8 the range of labeled application rates suggested for tank mixtures containing atrazine and alachlor from commercial product labeling. All agrochemicals were dissolved in 75mL methanol (MeOH), either individually or in combination, and then sprayed over the soil surface. To allow the MeOH to volatilize, bowls were placed in a fume hood overnight. The following morning, soils were rehydrated with 50mL spring water. The night prior to chemical exposure, all frogs were dehydrated in dry, unlined glass aquariums. After soil rehydration, frogs were immediately placed onto the contaminated soil surface for the 8-h exposure duration. Following exposure, 96 frogs were transitioned from their terrestrial to an aquatic microcosm for corticosterone collection as detailed below. After water samples were collected from the aquatic microcosms, frogs were immediately euthanized by submersion in liquid nitrogen, and brain tissues were excised for AChE analysis, followed by storage in a −80°C freezer until further processing. To evaluate tissue accumulation of herbicides, immediately following the 8-h exposure (i.e., without the 1-h depuration period required for corticoster one analyses), there maining frogs were euthanized by submersion in liquid nitrogen and stored in a −80°C freezer until extraction. Soil samples were pulled from each experimental unit and stored in a −80°C freezer for extraction and analysis.

Corticosterone analysis

To measure stress levels in frogs after agrochemical exposure, 1-h depuration water samples were obtained for each frog (n=12/treatment) following the methods of Gabor et al. (2013) for the waterborne assay. After exposure, frogs were transferred to clean Pyrex® bowls containing 100mL room temperature spring water. This water volume was enough to completely cover the entire body surface of each frog. After 1 h, water samples were collected and stored in a −80°C freezer prior to corticosterone analysis.

To analyse corticosterone levels (see Gabor et al. 2013), water samples were filtered through Waters® Oasis solid-phase extraction cartridges and eluted using MeOH; then the eluate was blown dry under nitrogen. Samples were then reconstituted and analyzed colorimetrically using Cayman Chemical Corticosterone ELISA Kits (96 wells) with an ELx808 Ultra Microplate Reader at 405nm (Biotek Instruments) coupled with Gen 5 2.04 software (Biotek Instruments). All samples were run in duplicate, and the final corticosterone values were averaged and standardized by individual amphibian body weight.

AChE analysis

Brain tissues were analyzed for AChE activity (n=12/ treatment) using methods modified from Ellman et al. (1961). Briefly, approximately 35mg of brain tissue was homogenized in a tris(hydroxymethyl)aminomethane (Tris):Triton buffer (1:5, w/v) and centrifuged at 4°C for 15min at 10000rpm. The resulting supernatant was kept on ice until used for final AChE analysis. Reaction mixtures contained a Tris:calcium chloride buffer, 5-5′-dithiobis-(2-nitrobenzoic acid), and acetylthiocholine iodide. All samples were run in duplicate and analyzed using a 10-min kinetic method on a Shimadzu UV1800 spectrophotometer set at 412nm.

Soil and amphibian agrochemical extraction

Pesticide extraction methods for amphibians and soils followed those outlined in Van Meter et al. (2014). Briefly, amphibians were weighed and tissues homogenized followed by freeze drying. Soil and amphibian samples were extracted 2 times in 5mL MeOH, blown down to approximately 1mL under nitrogen, and reconstituted with 10mL Milli-Q water. Three mL methyl-tert butyl ether (MTBE) was then added, and samples were allowed to gravimetrically separate before the addition of sodium sulfate. The MTBE layer was then pipetted off the surface, transferredtoa 2-mLcentrifugetube,and centrifugedat 13500rpm for 15min. The final sample was transferred to a gas chromatrography (GC) vial for analysis via GC–mass spectrometry (MS). The atrazine metabolites desethyl atrazine (DEA) and deisopropyl atrazine (DIA) were summed in total with atrazine when detected. From concentration data obtained for the frogs subject to the corticosterone treatment (n=12/treatment), bioconcentration factors (BCFs)were calculated for each frog as:

$$\text{BCF}=C_{\text{f}}/C_{\text{s}}$$

where C_f is the frog whole-body tissue concentration and C_s is the average composite soil concentration with in each treatment, both at the end of the 8-h exposure. Although BCFs typically refer to accumulation of contaminants from an aquatic medium at steady state, they also describe dietary and dermal accumulation in terrestrial environments, as presented in our study (Kenaga 1980; Henson-Ramsey et al. 2008).

To verify that urea concentrations were consistent, urea-N analysis was performed on control (C) and urea (N, ZN, LN, and ZLN) treated soils as associated with frogs in the corticosterone sample group only (n=12 samples/treatment; N =60). From each of these microcosms, 5-g soil samples were sent to the Agriculture Diagnostic Laboratory at the University of Arkansas (Fayetteville, AR, USA). Urea concentrations were determined using a KCl extraction assay on a BioTek microplate reader or a Skalar autoanalyzer following the methods of Huluka and Miller (2014).

GC-MS analysis

Soil and frog extracts were analyzed on an Agilent 7890A GC coupled to an Agilent Technologies 5975C MS. All data were collected and processed using ChemStation software. All injections (2mL) were made in splitless mode, and the carrier gas was helium maintained at a constant flow of 1.0mL/min. The inlet and transfer line were held constant at 275°C, and the MS source and MS quad temperatures were 230 and 150°C, respectively. For chromatographic separation, the initial oven temperature was held at 70°C for 1min, ramped 50°C/min to 150°C, ramped 6°C/min to 200°C, and finally ramped 16°C/min to 280°C and held for 4.0min (total runtime 19.9min). Atrazine, alachlor, and 2 metabolites of atrazine (DEA and DIA) were analyzed in selected ion monitoring mode using positive electron ionization. Atrazine was monitored at 200 and 174m/z, alachlor at 160 and 188m/z, DEA at 172 and 174m/z, and DIA at 173 and 158m/z ions. The internal standard, tetraconazole, was monitored at 336 and 338m/z ions. Blanks were run at the beginning and intermittently throughout the run to minimize carryover. Standards were analyzed at the start and end of each run, and quality assurance/quality control samples were analyzed throughout.

Statistical analysis

All analyses were performed in R Ver 2.11.1 (R Core Development Team 2015). Corticosterone (pg mL⁻¹ g⁻¹) and AChE(mol min⁻¹ mg⁻¹) were evaluated individually by pesticide treatment through analysis of variance (ANOVA) using the aov function. Post hoc analyses were performed using Tukey’s honestly significant difference function for corticosterone and the least significant difference function for the AChE analysis. The BCFs, calculated for atrazine and alachlor individually, were compared across treatments using ANOVA and Tukey’s honestly significant difference post hoc tests. When necessary, all data were natural log-transformed before analysis to adhere to assumptions of normality and homogeneity of variance.

RESULTS

Soils and amphibian tissues

Atrazine, alachlor, and urea concentrations in soils were consistent across treatment groups in which a particular agrochemical was applied (Table 1). Among control soils, pesticide concentrations were not detected or were below instrument detection limits. Accounting for these “blank” concentrations did not affect the resulting calculations of BCF and therefore they were omitted from statistical analysis.

TABLE 1:

Average agrochemical concentrations in soils and frog tissues (in ppm ± standard error)

	Soil			Tissue w/depuration		Tissue wo/depuration
Treatment	Atrazine	Alachlor	Urea-N	Atrazine	Alachlor	Atrazine	Alachlor
C	ND	ND	0.3±0.1	ND	ND	ND	ND
Z	12.7±1.1	NA	NA	0.6±0.1	NA	3.9±1.6	NA
L	NA	16.1±1.6	NA	NA	1.6±0.3	NA	2.3±1.0
N	NA	NA	1130.4±78.4	NA	NA	NA	NA
ZL	9.1±0.9	15.2±1.1	NA	0.6±0.2	2.6±0.7	3.5±1.3	4.6±1.6
ZN	8.6±0.3	NA	922.8±81.4	1.0±0.4	NA	1.3±0.5	NA
LN	NA	19.2±1.9	854.6±80.4	NA	1.7±0.3	NA	2.2±0.6
ZLN	10.4±1.2	15.2±1.6	931.8±62.3	1.5±0.5	2.9±0.8	3.7±1.4	4.4±2.0

C=control; Z=atrazine; L=alachlor; N=urea; ND=not detected or below instrument detecti detection limit; NA=not analyzed; Tissue w/depuration=frogs depurated for 1-h corticosterone collection; Tissue wo/depuration=frogs not depurated.

Frogs in the control treatment that were not exposed to an agrochemical had no measurable atrazine or alachlor in their tissues (Table 1). As expected, tissue concentrations among frogs that were euthanized immediately following the 8-h agrochemical exposure were greater than among frogs that were depurated for the 1-h corticosterone assay (Table 1). For the 6 frogs that were not depurated, atrazine concentrations were similar in the Z, ZL, and ZLN groups, although these concentrations were 63 to 67% higher than in the frogs in the ZN treated group (Table 1). Alachl or concentrations were greatest in the frogs receiving agrochemical treatments that also contained atrazine; ZL and ZLN (Table 1). Tissue concentrations of alachlor were 47 to 50% greater in the ZL- and ZLN-treated frogs relative to the L-treated frogs. For both atrazine and alachlor, tissue concentrations were lowest among frogs in the double chemical treatments that included urea (i.e., ZN and LN; Table 1).

Frogs that were subject to depuration for the 1-h corticosterone sample had measurable tissue concentrations of both atrazine (range=0.1–5.9ppm) and alachlor (range=0.9–6.6ppm), although measures were lower, likely attributable to metabolism and excretion processes occurring during the 1-h hormone collection period. At the end of the 1-h depuration period, frogs in the ZLN treatment had the highest concentrations, on average, of both atrazine and alachlor (Table 1). The atrazine concentration was 60% greater among the ZLN frogs relative to the Z and ZL frogs, and alachlor was 81% greater in ZLN frogs relative to frogs in the L and LN treatments. For alachlor, although the ZLN group had the highest tissue concentration on average, the concentration was only 10% greater than frogs in the ZL treatment, which also had high tissue levels after depuration. Interestingly, for frogs in the double chemical treatments that included urea (both ZN and LN),the frogs that were depurated had tissue concentrations that were 23% lower for both atrazine and alachlor relative to frogs in the same treatment that were not depurated.

BCF

The atrazine BCF for frogs after the 1-h corticosterone sampling period was significantly higher in treatments in which urea was present (N main effect ANOVA, p=0.021), although there was no effect of alachlor (L main effect ANOVA, p=0.272) or the combination of alachlor plus urea (L × N interactive effect ANOVA, p=0.633) on atrazine bioaccumulation. Frogs in the ZN and ZLN treatments had atrazine concentrations that were 65 and 70% greater than the individual Z treatment, respectively (Figure 1A). Alachlor BCF was not significantly altered by any of the agrochemical treatments among frogs that were depurated (Z main effect ANOVA, p=0.073; N main effect ANOVA, p=0.323; ZN interactive effect ANOVA, p=0.329). Nonetheless, alachlor concentration in the triple ZLN treatment was 50% greater and 20% greater in the ZL treatment relative to the individual alachlor treatment (Figure 1B), suggesting that there may be increased bioaccumulation of this herbicide when atrazine is present.

FIGURE 1:

Bioconcentration factor (BCF; mean ± standard error) of atrazine (A) and alachlor (B) in juvenile Southern leopard frogs (Lithobates sphenocephala) after exposure to individual or mixed agrochemicals and following a 1-h depuration period. Z = atrazine; L = alachlor; N = urea. Capital letters above bars represent statistically significant differences among treatments.

Corticosterone

The highest corticosterone levels were measured from frogs exposed to all 3 chemicals in the ZLN treatment (Figure 2).When paired with urea, atrazine increased corticosterone production by an average of 25% in the ZN treatment and 59% in the ZLN treatment relative to the individual atrazine treatment (Figure 2). Atrazine had a significant effect on corticosterone (pg mL⁻¹ g⁻¹) release among frogs tested in our study (Z main effect ANOVA, p=0.022) as well as a significant interactive effect when paired with urea (Z × N interactive effect ANOVA, p=0.005). However, there was no significant 3-way interactive effect (Z × L × N interactive effect ANOVA, p=0.374) on corticosterone. The combination of atrazine and alachlor also increased corticosterone production, on average, by 34% relative to the single Z treatment(Figure 2),but the interactive effect of these chemicals was only marginally significant (Z × L interactive effect ANOVA, p=0.051). Corticosterone measures were lowest, on average, among the control, atrazine, alachlor, and urea single chemical treatments as well as the LN double treatment (Figure 2). Alachlor, urea, and the combination of alachlor plus urea had no effect in altering corticosterone production (L main effect ANOVA, p=0.146; N main effect ANOVA, p=0.727; L × N interactive effect ANOVA, p=0.420, respectively).

FIGURE 2:

Corticosterone activity levels (mean ± standard error; pg mL⁻¹ g⁻¹) in Southern leopard frogs (Lithobates sphenocephala) following agrochemical exposure. C = control; Z = atrazine; L = alachlor; N = urea. Capital letters above bars represent statistically significant differences among treatments.

AChE

The 3 highest brain AChE concentrations were observed in frogs exposed to treatments that included atrazine (Z main effect ANOVA, p=0.007), that is, the Z, ZL, and ZN treatments (Figure 3). Frogs in the ZL treatment had the greatest average AChE concentrations, which were 35% greater compared with the control frogs and 40% greater than frogs in the double LN treatment. On average, urea decreased AChE activity by 20% relative to the atrazine treatments (urea main effect ANOVA, p=0.014). In addition to the control group, 3 of the groups containing urea (theN, LN, and ZLN treatments) had the lowest AChE concentrations among the frogs that were depurated for 1 h (Figure 3). Frogs in the N treatment group had AChE levels that were 37 and 26% lower than the Z and L treatments, respectively (Figure 3). No main or interactive effects were observed among frogs in the alachlor (main effect ANOVA, p=0.877) or combined agrochemical treatments (all interactive effects ANOVA, p>0.05).

FIGURE 3:

Acetylcholinesterase (AChE) activity (mean ± standard error; mol min⁻¹ mg⁻¹) measured from brain tissues in juvenile Southern leopard frogs (Lithobates sphenocephala) following agrochemical exposure. C = control; Z = atrazine; L = alachlor; N = urea. Capital letters above bars represent statistically significant differences among agrochemicals.

DISCUSSION

Agrochemical use and subsequent exposure to nontarget organisms is of concern for many species and life stages, notably amphibians (e.g., reviewed in Mannet al. 2009; Brühlet al. 2011; Van der Kraak et al. 2014; Pisa et al. 2015; van der Sluijs et al. 2015). Multiple agrochemicals have been detected simultaneously in field samples from residual and/or current use applications in terrestrial and wetland habitats where amphibians reside (Smalling et al. 2013, 2015; Christin et al. 2013; Glinski et al. 2018a; Swanson et al. 2018). To evaluate the sublethal impacts of agrochemicals, several biomarkers have been suggested as good candidates for evaluating exposures in natural populations (reviewed in Venturino and de D’Angelo 2005) including glutathione-S-transferase, AChE, and cytochrome P450 modulation. In the present study, higher pesticide BCFs, elevated corticosterone levels, and altered AChE activity in juvenile Southern leopard frogs after dermal contact with a common agrochemical mixture were observed.

Uptake of pesticides through dermal exposure in postmetamorphic amphibians is well documented in recent literature (Henson-Ramsey et al. 2008; Van Meter et al. 2014, 2015, 2016, 2018; Glinski et al. 2018b, 2018c). In the present study, immediately following the 8-h exposure, atrazine tissue concentrations were similar among frogs in all treatments, except for the combined treatments of atrazine with urea, in which atrazine concentrations were markedly lower. Alachlor tissue concentrations were unaffected by the presence of urea but were greatly increased by the presence of atrazine. This finding suggests that the presence of fertilizer has an inhibitory effect on atrazine uptake whereas atrazine accentuates the accumulation of alachlor. The same pattern of enhanced pesticide uptake when paired with atrazine was seen previously for metolachlor, malathion, and propiconazole in juvenile green frogs following dermal exposure on soils (Van Meter et al. 2018).

After the 1-h depuration period required for the nonlethal corticosterone sampling method used in the present study, the same patterns of tissue concentration and BCF were seen for alachlor: atrazine increased but urea decreased alachlor concentrations. However, atrazine BCF was greater after the depuration period in the treatments in which urea fertilizer was present. These results may indicate that following fertilizer exposure, frogs may metabolize and/or excrete certain pesticides, such as atrazine, more slowly than frogs exposed to pesticides but without simultaneous fertilizers. Van Meter et al. (2018) reported that changes in BCF among pesticide-exposed frogs were not consistent with alterations in the metabolome, and whereas BCF values may have seemed low, the biological impact they elicited may have been quite large. This enhanced retention of atrazine, although relatively low in terms of bioconcentration, may have serious consequences for amphibians given atrazine’s known impacts on endocrine, neurological, reproductive, and other essential functions (reviewed in Van der Kraak et al. 2014).

Corticosterone,the glucocorticoid stress hormonereleased in amphibians, plays a critical role in reproduction, immunefunction, metabolism, food acquisition, and behavior (e.g., Glennemaier and Denver 2001; Moore and Jessop 2003). As anticipated, Southern leopard frog juveniles exposed to the full agrochemical mixture of atrazine, alachlor, and urea (ZLN) had the highest waterborne corticosterone levels, as well as the greatest atrazine and alachlor BCFs (as discussed above), indicating a heightened stress response after this short-term exposure. Elevated corticosterone levels were also found in the frogs in the combined atrazine plus alachlor (ZL) and atrazine plus urea (ZN) treatments; however, the combination of alachlor plusurea (LN) resulted in the lowest corticosterone levels among all treatments tested, including the control group. These results suggest that the effects of atrazine on stress levels in juvenile amphibians may be magnified when atrazine is paired with one or more agrochemicals, as would realistically occur in a tank mixture. Briefly reviewed by Van der Kraak et al. (2014), mammalian research has shown that atrazine exposure increases blood corticosterone levels due to activation of the hypothalamic–pituitary axis. Similarly, plasma corticosterone levels were elevated in African clawed frog (Xenopus laevis) adults after exposure to a 9pesticideaquaticmixturethatincludedbothatrazineandalachlor (Hayes et al. 2006). However, because researchers did not examine the individual effects of the 9 pesticides tested or the paired effects of various chemical combinations leading up to the 9-pesticide mixture, it is difficult to evaluate the extent to which atrazineoralachlor played a significant role in the observed stress response. Nonetheless, our data corroborate those of Hayes et al. (2006), who reported elevated stress response in postmetamorphic amphibians following exposure to multiple, concurrent agrochemicals in the laboratory.

A field study measuring plasma corticosterone in bullfrogs (Lithobates catesbeianus) downstream of agricultural and nonagricultural sites found that frogs in the agriculturally impacted sites had elevated corticosterone levels relative to frogs from the nonagricultural sites (Falso 2011). Although water quality analyses revealed the presence of numerous pesticides and fertilizers, analyses were only performed on one agricultural and one nonagricultural site, making it difficult to draw broad conclusions about the relationship between the corticosterone levels measured and the presence of agrochemicals. In addition to elevated corticosterone levels, frogs from agricultural sites had elevated blood neutrophil, lymphocyte, eosinophil, and monocyte concentrations, indicating altered immune response (Falso 2011). In CostaRica, drab treefrogs (Smilisca sordida) were equally likely to bioaccumulate pesticides at low and high elevations, with no clear relationship to corticosterone or other hormones studied(Leary et al. 2018).Future studies that make use of a broader range of nonlethal biomarkers following agro chemical exposure, such as advances in skins wabbing and urine analysis in amphibians for corticosterone evaluation (Santymire et al. 2018; Narayan 2013, respectively) are strongly encouraged.

Interestingly, in agricultural areas, factors in addition to agrochemical presence can alter corticosterone levels. In a substrate experiment, adult toads (Bufo bufo) had elevated corticosterone levels when exposed to bare, ploughed soils relativeto soils covered with leaf litter or grasses, and individuals actively avoided bare soils in a choice experiment (Janin et al. 2012). Juvenile toads tended to prefer bare soils and there was no stress response associated with this substrate choice. Therefore, this may leave juveniles more vulnerable to agrochemical exposures on bare soils. In the context of amphibian movement across agricultural landscapes, stress associated with ploughed soils may be compounded by stress striggered through chemical exposure.

Agrochemical exposure may also alter the proper neurotransmitter function that is essential in amphibians for predator avoidance, mating behavior, and prey capture. The enzyme AChE has been used as a biomarker for agrochemical and metal exposure among amphibians in many field and laboratory studies because it functions to stop transmission of nervous impulses between synapses (reviewed in Venturino and de D’Angelo 2005; Mann et al. 2009). In the present study, atrazine increased AChE levels, whereas urea decreased AChE activity in juvenile Southern leopard frogs. The greatest AChE activity levels were observed among frogs in the atrazine plus alachlor treatment, although frogs in the individual atrazine, alachlor, and atrazine plus urea treatments had similarly high levels of AChE activity. Van Meter et al. (2018) reported upregulation of tyrosine among frogs exposed to mixed pesticide treatments including atrazine, dichlorophenoxyacetic acid, and metolachlor, which may also be indicative of changes to neurotransmitter function that result in neural overstimulation and coincides with the increases in AChE activity seen in the present study.

Alterations to AChE activity following agrochemical exposure in postmetamorphic amphibians reported in other laboratory and field studies suggest that AChE response varies across many biological, chemical, and environmental factors. In bullfrogs (L. catesbeianus) studied in the field in Canada, significantly higher AChE levels were found as agricultural intensity increased, with corresponding increases in atrazine levels (Marcogliese et al. 2009). Greater parasite loads were also seen along the agricultural gradient, which may have contributed to the elevated AChE levels reported. The opposite pattern of decreased plasma AChE levels was reported in wild-caught Indian bullfrogs (Hoplobatrachus tigrina) and Asian common toads (Duttaphrynus melanostictus) following intraperitoneal exposure to methyl parathion or carbaryl (Kumari and Sinha 2009). Although brain AChE was also inhibited among parathion- and carbaryl-treated toads relative to control frogs, brain AChE was only significantly lower than the control for carbaryl-treated frogs. Concurrent with alterations in AChE, the authors also reported muscle twitching, convulsions, loss of coordination, paralysis, difficulty breathing, nasal secretions, and discoloration of the limbs for both carbaryl- and parathiontreated frogs. In the Ukraine, adult marsh frogs (Rana ridibunda) collected from reference and industrialized urban sites and subsequently exposed to the fungicide procamocarb for 14 d in the laboratory showed alternate trends in AChE activity (Falfushinska et al. 2008). Frogs from the reference site had increased AChE and those from the urban site had decreased AChE following pesticide exposure. Previous exposure to contaminants combined with additional environmental stressors may make amphibians more susceptible to agrochemical compounds, but assessing neurotransmitter functionality following exposure may be difficult.

A field study exploring cholinesterase enzyme levels among amphibians living in rice paddies where agrochemical use is prevalent provides further evidence that environmental exposure to agrochemicals has deleterious effects. In Western Ghats, India, common frogs (Fejarvarya limnocharis) were collected from rice paddies throughout a 5-mo period that coincided with pesticide and fertilizer applications (Hedge and Krishnamurthey 2014). Liver and brain AChE activity was highest among frogs collected from the reference sites and was lowest among frogs living in the intermediate and highest contaminated sites, respectively. Interestingly, when AChE was paired with urea in our study, AChE levels were consistently lower relative to the corresponding individual atrazine and alachlor or double herbicide (ZL) treatments. Because we were unable to quantify urea concentrations in frog tissues, we are unable to relate urea bioaccumulation patterns directly with measured effects. Nonetheless, lower AChE activity levels reported, coupled with the field study presented previously (Hedge and Krishnamurthey 2014) and reviewed by Mann et al. (2009), suggest that fertilizer exposure may inhibit neurotransmitter activity, increase mortality, and alter behavior in frogs.

Data on pesticide concentrations and associated effects in amphibians from field exposures are limited, but greatly needed. Swanson et al. (2018) used passive sampling devices paired with radio telemetry of Northern leopard frogs (Lithobates pipiens) to monitor pesticide loads in habitats that mirrored amphibian movements in an agriculturally intensive landscape. They discovered that Northern leopard frogs were only found in an agricultural habitat approximately 6% of the time, but that this habitat had the greatest pesticide loads. Therefore, even limited use of agricultural habitats by amphibians may present a great risk for pesticide exposure. Smalling et al. (2015) also reported pesticide loads in wetland water and associated bed sediments, along with amphibian body burdens. The pesticide loads reported for amphibians in these 2 field studies were much lower than those we report in the present study (ppb vs ppm, respectively). The pesticides used in the present study were applied at the maximum application rate, which represents a worst-case scenario experiment in which frogs are exposed on bare soils immediately following agrochemical application. Inherent differences between acute laboratory studies and chronic field exposures make it challenging to predict pesticide effects in amphibians in agricultural settings. Future study designs that work to bridge the knowledge gap between pesticide exposure and related effects, by including a broader range of pesticide concentrations and exposure scenarios, are strongly encouraged.

Co-exposure to multiple agrochemicals can elicit more pronounced adverse effects among juvenile amphibians, as seen in the present study and supported by data from Van Meter et al. (2018). Further testing of common tank mixtures, including additional fertilizing compounds, on terrestrial phase amphibians is prudent. Furthermore, acute studies should be compared with chronic exposures both in the laboratory and in the field to better understand how well these biomarkers predict real-world exposure scenarios and ultimately alter amphibian populations. Our current lack of knowledge with respect to pesticide mode of action in nontarget organisms presents many challenges for risk assessment; however, continued use of numerous biomarkers following agrochemical exposure is essential in building a strong foundation to better inform regulatory practices.