The Fungicide Chlorothalonil Is Nonlinearly Associated with Corticosterone Levels, Immunity, and Mortality in Amphibians

Background: Contaminants have been implicated in declines of amphibians, a taxon with vital systems similar to those of humans. However, many chemicals have not been thoroughly tested on amphibians or do not directly kill them. Objective: Our goal in this study was to quantify amphibian responses to chlorothalonil, the most commonly used synthetic fungicide in the United States. Methods: We reared Rana sphenocephala (southern leopard frog) and Osteopilus septentrionalis (Cuban treefrog) in outdoor mesocosms with or without 1 time (1×) and 2 times (2×) the expected environmental concentration (EEC) of chlorothalonil (~ 164 μg/L). We also conducted two dose–response experiments on O. septentrionalis, Hyla squirella (squirrel treefrog), Hyla cinerea (green treefrog), and R. sphenocephala and evaluated the effects of chlorothalonil on the stress hormone corticosterone. Results: For both species in the mesocosm experiment, the 1× and 2× EEC treatments were associated with > 87% and 100% mortality, respectively. In the laboratory experiments, the approximate EEC caused 100% mortality of all species within 24 hr; 82 μg/L killed 100% of R. sphenocephala, and 0.0164 μg/L caused significant tadpole mortality of R. sphenocephala and H. cinerea. Three species  showed a nonmonotonic dose response, with low and high concentrations causing significantly greater mortality than did intermediate concentrations or control treatments. For O. septentrionalis, corticosterone exhibited a similar nonmonotonic dose response and chlorothalonil concentration was inversely associated with liver tissue and immune cell densities (< 16.4 μg/L). Conclusions: Chlorothalonil killed nearly every amphibian at the approximate EEC; at concentrations to which humans are commonly exposed, it increased mortality and was associated with elevated corticosterone levels and changes in immune cells. Future studies should directly quantify the effects of chlorothalonil on amphibian populations and human health.


Research
Amphibians are arguably the "poster child" of the present extinction crisis (Wake and Vredenburg 2008), with > 32% of species threatened and at least 43% experiencing popu lation declines (Stuart et al. 2004). Chemical pollution is a concern for the health of both amphibians and humans. It is consid ered the second greatest threat (behind habitat loss) to aquatic and amphibious species in the United States and has been linked to amphib ian declines and disease (Davidson et al. 2002;Rohr et al. 2008a). Similarly, contaminants have been linked to mortality and disease in humans (Dietert et al. 2010). Importantly, many vital systems of amphibians, such as endocrine and immune systems, are simi lar to those in humans (Hayes 2005), and a genome analysis revealed that the amphib ian Xenopus tropicalis has > 1,700 genes with human disease associations (Hellsten et al. 2010). Thus, in addition to being of con servation concern, amphibians might be an under used model taxon for studying stressor effects on human health.
Although the hypothesis that contaminants are a factor in amphibian declines is plausible, most previously tested chemicals have not directly killed amphibians at or below expected environmental concentrations (EECs; but see Rohr et al. 2006b;Storrs and Kiesecker 2004), although sub lethal and indirect effects can be strong (Rohr et al. 2006a). Nevertheless, many chemicals remain untested on amphib ians. For example, chloro thalonil is the most commonly used synthetic fungicide in the United States (Kiely et al. 2004) and is toxic to shrimp, insects, and fish at or below the EEC (164 μg/L) (Caux et al. 1996;Grabusky et al. 2004). It can be transported great distances and has been found in tropical mountains where most amphibian declines have occurred (Stuart et al. 2004). However, its effects on amphibians have rarely been studied.
Chloro thalonil can also affect vertebrate and invertebrate immune systems. Chloro thalonil exposure was associated with con tact dermatitis (Penagos 2002) and DNA damage to leukocytes of farmers 1 day after spraying (Lebailly et al. 1997). It can be immuno suppressive to oysters and fish, reducing macrophage viability and function, immuno globulin M, and expression of pro inflammatory cyto kines (BaierAnderson and Anderson 2000; Shelley et al. 2009). These findings are a concern because pollution is often associated with wildlife disease emer gence (Dobson and Foufopoulos 2001) and amphibians are being decimated by infectious disease (Daszak et al. 2003). The objective of this study was to quantify the effects of chloro thalonil on amphibian survival, immu nity, corti costerone levels, and liver density.
Chloro thalonil (2,4,5,6tetrachloro iso phthalo nitrile) is widely used to control fungus on peanuts, corn, and potatoes (Cox 1997). Approximately 14 million pounds are applied annually in the United States, with approximately 500,000 pounds used per year in Florida [U.S. Environmental Protection Agency (EPA) 1999], the location of the pres ent study. Chloro thalonil is typically applied during the wet season, corresponding to the breeding activity of many amphibians (Rohr et al. 2004).
Chloro thalonil binds to glutathione, which disrupts cellular respiration (Grabusky et al. 2004), a vital process for virtually every organism, including humans. In water, chloro thalonil is short lived, with a halflife of approximately 44 hr (U.S. EPA 1999). The primary chloro thalonil metabolite (4hydroxy 2,5,6trichloro iso phthalo nitrile) is estimated to be 30 times more acutely toxic than chloro thalonil and is also more persistent and mobile in soil (U.S. EPA 1988). During its manu facture, chloro thalonil is also contaminated with hexachloro benzene (Hung et al. 2010), a probable carcinogen with a soil halflife of 3-6 years (Cox 1997). Shuman et al. (2000) detected chloro thalonil concentrations of ≤ 290 μg/L in run off, and chloro thalonil has been detected in ground water ("standpipe" wells) at concen trations ≤ 272 μg/L. Nevertheless, the EEC of chloro thalonil in ponds [calculated using the U.S. EPA's GENEEC software, version 2; for parameters, see Supplemental Material,

Materials and Methods
This work was approved by animal care and use committees of the University of South Florida (W3228) and the University of Florida (02308WEC). All animals used were treated humanely and with regard for alleviation of suffering.
Mesocosm experiment. Tanks were arranged in a randomized block design with four replicates of each treat ment (a total of 16 tanks). There were two control treatments, receiving either 50 mL of water or 50 mL acetone solvent (used to dissolve chloro thalonil). Tanks for the remain ing two treatments received chloro thalonil (technical grade, purity > 98%; Chemservice, West Chester, PA) dissolved in 50 mL acetone so that nominal concentrations in the tanks were either one time the EEC (1×; 164 μg/L) or two times the EEC (2×; 328 μg/L). Tanks were dosed the same day as the amphibians were added, and targeted nominal concen trations closely matched the actual concen trations (1×, 172 μg/L; 2×, 351 μg/L; spiked recovery efficiencies, 95%). Thus, for simplic ity and consistency across the experi ments in this article, we refer to the nominal concentra tions. Several water quality and chemis try vari ables were quantified at various times during the experi ment [see Supplemental Material, "Mesocosm Experimental Methods" and Tables S3 andS4 (doi:10.1289/ehp.1002956)]. Standardized dip net sampling of each tank was conducted the third day of the experi ment to quantify any rapid mortality associated with chloro thalonil exposure. The number of metamorphosed frogs was noted daily, and tadpole survival was determined 5 weeks after dosing.
Laboratory experiment I. We obtained Hyla squirella and O. septentrionalis from multiple, thoroughly mixed clutches col lected from two adjacent retention ponds in Tampa, Florida, in July 2008 (N 28°0.322´, W 82°19.532´). We employed a completely randomized design with 21 32L glass aquaria, each filled with 10 L artificial spring water (Cohen et al. 1980), with water hardness of 62.7 ppm (5B Hardness Test Kit; HACH Co., Loveland, CO) and pH ~ 7.0). Aquaria were maintained in a laboratory at the University of South Florida at 27°C and on a 14:10hr light:dark cycle. Each aquarium received five H. squirella and 15 O. septentrionalis tadpoles , and tadpoles were fed boiled organic spinach daily. We used five treatments of technical grade chloro thalonil (purity > 98%; Chemservice; actual concentra tions, 0.176, 1.76, 17.6, 176, and 1,760 μg/L) and two control treatments [water and solvent (500 ng/L acetone)], with three replicates per treatment. The targeted nominal concentration for the chloro thalonil stock was 1,640 μg/L, and the actual concentration was 1,760 μg/L (spiked recovery efficiencies, 95%). All of the other concentrations were attained through serial dilutions of this stock solution. Again, for simplicity and consistency across the experi ments, we refer to the nominal concentrations. We quantified frog survival and preserved dead tadpoles 12 hr after the start of the experiment and then every 24 hr for 4 days (there were no water changes). Surviving tadpoles were eutha nized and preserved (70% ethanol) at the end of the experiment.
Laboratory experiment II. The same pro tocols used in laboratory experiment I were used in this experiment, conducted in October 2008, with the following exceptions. We tested three tadpole species: R. sphenocephala, O. septentrionalis, and H. cinerea (all starting at Gosner stage 25). We employed a com pletely randomized design with 144 500mL mason jars, each filled with 300 mL artificial spring water and each receiving three tadpoles of a single species. Species were isolated in this experiment because O. septentrionalis was occasionally observed depredating H. squirella in laboratory experiment I. The jars received one of six chloro thalonil treatments (0.0164, 0.164, 1.64, 16.4, 82.0, or 164 μg/L) or water or solvent. We used the same stock solution as in laboratory experiment I. A single water change occurred on day 7, and each jar was redosed at that time. There were six replicates per species per treatment. The number of sur viving tadpoles was noted after 4 hr, 24 hr, and then every 24 hr, for 10 days, and all dead tad poles were removed and preserved in formalin at those times.
To quantify the effects of chloro thalonil on tadpole livers and immune cells, at the end of the experiment one arbitrarily selected O. septentrionalis from each replicate was euthanized, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. We used O. septentrionalis for liver, immune, and corti costerone quantification because it had the lowest mortality of the three species and thus offered us the most survivors per tissue. To test whether chloro thalonil expo sure affected liver tissue integrity, we used ImageJ64 software (Rasband 2010) to cal culate liver tissue density, following ImageJ's Quantifying Stained Liver Tissue (Burger and Burge 2009), which reports the density of stained tissue within a designated area. To test whether chloro thalonil exposure affected density of liver immune cells, we counted the number of melano macrophages and granulo cytes per field of view at 400× magnification. Melanomacrophages and granulo cytes are leuko cytes that help defend against a variety of parasites (Rohr et al. 2008b). Because of the morphological similarity among granulocytes, we conservatively categorized all granule containing immune cells as granulo cytes, but most were likely eosinophils.
Corti costerone experiment. We used O. septentrionalis tadpoles (Gosner stages 25-28; the same population as used in labora tory experiment II) to quantify the effect of chloro thalonil exposure on frog corti costerone levels, a steroid hormone elevated in response to natural and anthropogenic stressors, includ ing pesticides (Martin et al. 2010). We used the same general protocols as described in laboratory experiment II and the following treatments: 0.164, 16.4, 82, and 164 μg/L chloro thalonil, and water and solvent con trols. These treatments were crossed with one of three chloro thalonil exposure durations: 4, 28, or 100 hr (n = 3, 2, and 3, respectively). The exception, however, was that tadpoles exposed to 164 μg/L chloro thalonil were only exposed for 4 hr because they did not sur vive for 28 or 100 hr of exposure. This design resulted in 43 independent replicates. After the appropriate exposure duration, tadpoles were euthanized, and individual tadpoles were weighed (to 0.0001 g) and homogenized in ultrapure water. Tritiated corti costerone (2,000 cpm) was then added to each sample to quantify recoveries post extraction. We used a corti costerone enzyme immuno assay (EIA) kit (catalog no. 900097; Assay Designs, Ann Arbor, MI) to quantify hormone levels in each sample. Individual recoveries (mean, 55.3%) and tadpole mass measurements were used to estimate corti costerone per gram of tadpole tissue. Detailed methods for this EIA kit and a discussion of its potential limita tions are provided in Supplemental Material (doi:10.1289/ehp.1002956).
volume 119 | number 8 | August 2011 • Environmental Health Perspectives Statistical analyses. For all experiments and responses, we compared the water and solvent controls. Because we found no differ ence between these treatments (p > 0.328), we pooled the two treatments into one "control" group for all subsequent analyses.
For the mesocosm experiment, all analyses were conducted on the arcsinesquareroottransformed proportion of R. sphenocephala and O. septentrionalis surviving to the end of the experiment, controlling for the four spatial blocks. We tested whether chloro thalonil was associated with mortality rela tive to the control treatments by conducting a permutationbased multi variate regression analysis. For the laboratory experiment, we arcsinesquareroot transformed the propor tion of tadpoles surviving until the end of the experi ment and log transformed hours to death, mass of survivors, amount of liver damage, and melanomacrophage and granu locyte counts to meet parametric assumptions. For the liver and immune analyses, we log transformed chloro thalonil concentration and weighted the time to death analyses by the number of animals that died per replicate. If a dose response appeared linear, chloro thalonil concentration was treated as a continuous predictor in a regression model (liver density). If a dose response was non linear but relatively simple (one inflection point), chloro thalonil concentration was treated as a continuous predictor, and we used polynomial regres sion with type II sums of squares to fit the data (immune responses). If a response was non linear and relatively complex (more than one apparent inflection point), chloro thalonil concentrations were treated as levels of a cat egorical predictor followed by Fisher's least significant difference (LSD) multiple com parison test to determine which levels were different from one another (proportion of tadpoles that survived and time to death). As an additional test for non monotonicity (humpshaped dose response), we eliminated the highest concentrations, which typically caused considerable mortality, and used poly nomial regression to test for a quadratic doseresponse relation ship with the remaining concentrations. For the immune responses, we conducted a multi variate polynomial regression model with melano macrophages and granulocytes as responses and followed it by uni variate analy ses on each response vari able. We loglog transformed these relation ships to improve fit and meet the assumption of the polynomial regression.
For the corti costerone experiment, we conducted polynomial regression (using least trimmed squares) with chloro thalonil con centration as a continuous predictor and log transformed corti costerone as the response vari able. All statistical analyses were conducted with Statistica (version 8.0; Statsoft, Tulsa, OK). We did not calculate LC 50 (concen tration that results in death of 50% of individuals by a given time) values for any responses because all three dose-response experiments showed evidence of non monotonicity, which would violate the assumptions of LC 50 calculations.
A mean (± SE) of 1.5 ± 0.327 live tadpoles were captured per dip netting session in control tanks, but no live tadpoles were netted from chloro thalonil tanks (the only tanks where dead tadpoles were netted). These results sug gest that most of the mortality associated with chloro thalonil occurred within the first 72 hr of exposure.
Laboratory experiment I. Survival was lower for H. squirella than for O. septentri onalis, most likely because O. septentrionalis depredated H. squirella (Figure 2A). Despite this predation, time to death for H. squire lla was shorter for tadpoles exposed to any chloro thalonil concentration relative to con trols [Fisher's LSD, p < 0.023 for controls compared with any chloro thalonil concentra tion ( Figure 2B); for full analysis of cova riance results, see Supplemental Material (doi:10.1289/ehp.1002956)].
For O. septentrionalis, survival was non monotonic, with low and high concentrations causing significantly greater mortality than intermediate concentrations and control treat ment (Figure 2A). Relative to controls, survival was reduced by > 80% in the 0.164, 17.6, 164, and 1,640 μg/L concentrations, but survival was not significantly reduced by 1.64 μg/L chloro thalonil, and this concentration was sig nificantly different from both adjacent concen trations (Figure 2A). This non monotonicity was also supported by polynomial regression, which produced a significant quadratic term for concentrations < 16.4 μg/L [for statistics, see Supplemental Material (doi:10.1289/ ehp.1002956)]. Relative to controls, time to death was shorter for O. septentrionalis tad poles exposed to any chloro thalonil concentra tion (Fisher's LSD, p < 0.021 for 0 μg/L vs. 0.164, 1.64, 164, or 1,640 μg/L; Figure 2B), with the exception of 16.4 μg/L (Fisher's LSD, p = 0.190; Figure 2B).

Laboratory experiment II.
For each spe cies, the 164 μg/L concentration killed 100% of the tadpoles by the end of the experi ment [ Figure 2C; for mortality through time and full statistical results, see Supplemental Material, Figure S1 and methods for labora tory experiment II, respectively (doi:10.1289/ ehp.1002956)]. However, we observed con siderable variation among species in their sensitivity to chloro thalonil. R. sphenocephala appeared most sensitive, experiencing 86% mortality at 0.164 μg/L and 100% mortality at 82 μg/L ( Figure 2C), whereas O. septentri onalis was least sensitive (Figure 2A).
The dose response for survival was signifi cantly non monotonic for R. sphenocephala and H. cinerea, with low and high concentra tions causing significantly greater mortality than intermediate concentrations and con trol treatment ( Figure 2C), a result similar to the non monotonic dose response revealed in laboratory experiment I for O. septentrionalis. For R. sphenocephala, 0.164 μg/L caused sig nificantly more mortality than did each adja cent concentration, and we found a signifi cant quadratic term for the response to doses < 82 μg/L. For H. cinerea, 0.0164 μg/L caused significantly more mortality than did each adjacent concentration, and as for R. sphe nocephala, we found a signifi cant quadratic term for the response to doses < 16.4 μg/L [for poly nomial results for both species; see Supplemental Material (doi:10.1289/ ehp.1002956)]. As a reminder, each data point in Figure 2C is the mean of six data points, and thus the 0.0164 μg/L concentration for H. cinerea is not an outlier or artifact.
O. septentrionalis did not exhibit a non monotonic response in this experiment as it did in laboratory experiment I (Figure 2A,C). This is possibly due to differences in tadpole densities, developmental stages, source popu lations, or bio accumulation of chloro thalonil associated with O. septentrionalis depredat ing H. squirella in laboratory experiment I. Chloro thalonil has been documented to bioaccumu late up to 3,000 times in fish (Cox 1997;U.S. EPA 1999).
Increasing chloro thalonil concentrations were associated with significant decreases in liver density of O. septentrionalis [F 1,40 = 4.82, p = 0.034; Figure 3A; see also Supplemental Material, Figure S2 (doi:10.1289/ehp. 1002956)]. Chloro thalonil concentration was also associated quadratically with both melano macrophages and granulocytes in this species ( Figure 3B; for statistics, see Supplemental Material). That is, relative to controls, tadpoles exposed to low concentrations had fewer of these immune cells, whereas tadpoles exposed to high concentrations had elevated numbers of these cells ( Figure 3B). We observed con siderable mortality at the 82 and 164 μg/L concentrations that may have confounded our immune results and might explain the increase in immune cells at these concentra tions. Thus, we reanalyzed the dose response excluding these two highest concentrations and discovered that, at these lower and more environmentally common concentrations, chloro thalonil was associated with a reduc tion in both melanomacrophages (F 1,32 = 4.67; p = 0.038) and granulocytes (F 1,32 = 5.52; p = 0.025; Figure 3B).
Corti costerone experiment. Corti costerone levels increased significantly with chloro thalonil exposure duration (F 1,27 = 11.57, p = 0.002). After 4 hr exposure to chloro thalonil, the relationship between log corti costerone levels and chloro thalonil con centration was significantly nonlinear (concen tration 3 : F 1,11 = 6.12; p = 0.031), with low and high concentrations of chloro thalonil being associated with higher levels of corti costerone than were intermediate concentrations and controls ( Figure 4). Multiple comparison tests further supported the conclusion that this dose-response curve was significantly nonlinear (Figure 4). This same general pat tern persisted for up to 100 hr of exposure, but tadpoles were not available after the 4 hr exposure duration for 164 μg/L because of high mortality (Figure 4). As a reminder, we conducted this study on the O. septentrionalis population that did not exhibit any significant non monotonic mortality response to chloro thalonil and exhibited significant mortality only at concentrations ≥ 82 μg/L ( Figure 2C).

Discussion
Ultimately, scientists should use a weightof evidence approach to evaluate risk posed by chemicals, which is partly why we conducted four experiments to quantify the effects of chloro thalonil on amphibians: a contrived, but highly controlled, laboratory experiment (labora tory experiment II), a more ecologi cally rele vant laboratory experiment where we allowed species to inter act (laboratory experi ment I), a laboratory experiment to assess whether corti costerone levels exhibited a dose response similar to that for mortality (corti costerone experiment), and a field meso cosm experiment with a complex freshwater community (mesocosm experiment). In all of these experiments, we found adverse effects of chloro thalonil on tadpoles. Although in laboratory experiment I we had low survival of H. squirella in the control group, possibly because of depredation by O. septentrionalis, these species regularly coexist, making this inter action ecologically relevant. This experi ment also reinforced the significant lethal ity of the EEC and lower concentrations of  Figure 4. Effects of chloro thalonil on corti costerone per gram of O. septentrionalis tissue shown as least squares means ± 1 SE. Means were averaged across the three chloro thalonil exposure durations (4, 28, and 100 hr), except for the 164 μg/L concen tration, where only the 4 hr duration mean is shown because longer exposure killed the tadpoles. Also shown is the significant thirdorder polynomial function (y = 1.886571 + 0.035582x -0.000668x 2 + 0.000003x 3 ) for the relation ship between chloro thalonil concentration and log corti costerone, adjusted for the effect of exposure duration. The corti costerone level for the 164 μg/L concentra tion is under estimated because it is the only mean based on 4 hr, rather than an average of 44 hr, of chloro thalonil exposure, and corti costerone increased significantly and loglinearly with the duration of chloro thalonil exposure (coefficient for log exposure duration = 0.269). Concentrations with different lower case letters are significantly different from one another by Fisher's LSD multiple comparison test (n = 13, 5, 7, 6, and 2 for 0, 0.164, 16.4, 82.0, and 164 μg/L, respectively).  (Figure 2A). We conducted a followup experiment using three amphibian species, this time preventing hetero specific inter actions. This experiment had 80-100% survival of the control tadpoles, simplifying data interpretation. It revealed that all three species were highly susceptible to chloro thalonil, with the EEC causing 100% mortality of each species in < 10 hr of expo sure. Moreover, in this experiment, we found evidence of non monotonic dose responses for mortality and fullbody measure ments of corti costerone, with low and high levels elevat ing both responses. Finally, in our mesocosm study, both the 164 and 328 μg/L concentra tions significantly reduced amphibian survival, suggesting that the labora tory results might be relevant to effects in nature. Together, these four experiments indicate that amphibians, in general, are susceptible to the EEC of chloro thalonil and that even low concentrations can cause amphibian mortality and physiological stress responses.
Our finding that amphibians are sensi tive to chloro thalonil is consistent with studies examining the sensitivity of aquatic vertebrates and invertebrates to chloro thalonil. For sev eral fish species, 48 and 96hr LC 50 values are < 20 μg/L and LOECs are near 1 μg/L chloro thalonil (Caux et al. 1996). The 48hr LC 50 for Bufo bufo japonicas was 160 μg/L (Hashimoto and Nishiuchi 1981). Daphnia magna had delayed reproduction when exposed to 32 μg/L (Ernst et al. 1991); in fathead minnows ≥ 6.5 μg/L chloro thalonil decreased the num ber of eggs per spawn, egg hatchability, and fry survival (as cited by Grabusky et al. 2004). The LOEC for H. cinerea and R. sphenocephala survival in our study was 10,000 times less than the EEC (0.0164 μg/L; Figure 2C) and was the lowest concentration we tested. Hence, we did not test low enough concentrations to detect a no observable effect concentration for the survival of these two species.
Three of the four amphibian species that we tested showed evidence of a non monotonic dose-mortality response to chloro thalonil (O. septentrionalis, Figure 2A; H. cinerea and R. sphenocephala, Figure 2C), with low and high levels causing significantly greater mortality than did intermediate levels and controls. Furthermore, for all species and experiments, the lowdose increase in mortal ity occurred within a single order of magni tude (either 0.016 or 0.16 μg/L). Although the non monotonic dose response for sur vival was observed for O. septentrionalis in only one of the two experiments (labora tory experi ment I; these experi ments used differ ent conditions and source populations), in the experi ment where O. septentrionalis did not exhibit a non monotonic dose response for survival (labora tory experi ment II), it did exhibit a non monotonic dose response for corti costerone. Hence, the non monotonic response was consistent and reproducible both within and across species, but whether low dose exposure to chloro thalonil and the associ ated stress response cause mortality appears to be context dependent. Non monotonic responses are important because they defy the traditional toxicological assumption that higher concentrations of a contaminant always cause greater harm. Nonmonotonic patterns have been observed previously in response to chloro thalonil (Shelley et al. 2009) and other agro chemicals (Storrs and Kiesecker 2004). Non monotonic responses can be caused by multiple mechanisms, affecting responses dif ferently at different doses, or by endocrine disruption (Welshons et al. 2003). Indeed, the Canadian Wildlife Service concluded that chloro thalonil might qualify as an endocrine disruptor because it has the potential to inter fere with endogenous hormones and enzymes and is an immuno modulator (Grabusky et al. 2004). However, the mechanism or mechanisms underlying non monotonic dose responses in this study remain unknown.
In addition to mortality, chloro thalonil was associated with immuno modulation of the surviving O. septentrionalis tadpoles. This finding is consistent with DNA damage to mono nuclear leukocytes of farmers 1 day after spraying chloro thalonil (Lebailly et al. 1997) and with studies of chloro thalonilinduced immuno suppression of fish and marine inver tebrates (BaierAnderson and Anderson 2000). Increases in chloro thalonil concentration up to 17.6 μg/L, concentrations to which humans are commonly exposed (Daly et al. 2007), were associated with reduced liver granulo cytes and melano macrophages in tadpoles, whereas further increases in chloro thalonil caused increased numbers of liver granulocytes and melano macrophages ( Figure 3B). This increase in immune cells might be in response to chloro thalonilinduced liver damage, based on our observations of decreased O. septentrio nalis liver density at these higher concentra tions [see Supplemental Material, Figure S2 (doi:10.1289/ehp.1002956)]. Alternatively, the increase in immune cells might itself have contributed to liver damage, because high lev els of melano macrophages and granulo cytes can cause tissue damage (Rose et al. 1999).
Although not yet studied, it is possible that exposure to chloro thalonil could reduce toler ance and resistance to parasites, which has been shown for wildlife and humans exposed to other agrochemicals (Dietert et al. 2010;Rohr et al. 2008b). If so, this could further reduce tadpole survivorship.
To our knowledge, we provide the first evidence that chloro thalonil elevates corti costerone. The significant non monotonic dose response of corti costerone to chloro thalonil was qualitatively similar to the mortality responses we observed in this study, underlin ing the consistent presence of non monotonic responses to this chemical. However, we do not know the direction of causation. Approaching mortality could have resulted in a systemic stress response that altered corti costerone and immune parameters; changes in corti costerone and immune parameters could have caused the mortality; or both of these scenarios could have occurred. Mortality at the highest concentrations of chloro thalonil seemed to occur too quickly to be mediated by corti costerone. However, it is plausible that corti costerone could have been involved in the mortality and immune cell changes observed at low concentrations of chloro thalonil. First, corti costerone is known to cause elevations in circulating granulo cytes in other animals (Davis et al. 2008), either by inducing proliferation or by efflux from cell reservoirs. Second, continuously elevated corti costerone has mani fold negative effects on health, including muscle atrophy, reduced neuro genesis, and immune suppression or dys regula tion (Martin 2009). Lastly, gluco corticoids, including corti costerone, are com monly elevated in response to stressors, natural and anthropogenic (Martin et al. 2010), and even in cases where elevations are insufficient to cause mortality, they can generally com promise health, even in humans (Wingfield and Sapolsky 2003). Although we cannot say with certainty whether the immuno logical effects observed in this study were mediated by corti costerone, we strongly advocate future efforts to assess the role of chloro thalonil and glucocorticoids as potential endocrine disrup tors, especially as disruptors of the immune system and disease resistance.
Although pesticides have been suggested as a cause of amphibian declines, there are few convincing cases in which pesticides cause high enough mortality at environmen tally realistic concentrations to directly affect amphibian populations (Belden et al. 2010;Rohr et al. 2006b). Sometimes even high mortality of larval amphibians can have little observable effect on the population because of densitymediated compensation, in which survivors of a factor experience lower mortal ity than do control animals after the stres sor is removed because of less competition for resources (Rohr et al. 2006b). However, based on amphibian demographic models that incorporate negative density depen dence (Vonesh and De la Cruz 2002), the level of EECinduced mortality reported here would likely reduce amphibian population sizes. Given that chloro thalonil caused nearly 100% mortality at the EEC, caused signifi cant mortality at concentrations four orders of magnitude below the EEC, and caused immuno modulation in surviving individuals, exposure to this chemical has the potential to both directly and indirectly cause amphib ian declines. Indeed, frog dieoffs have been reported after chloro thalonil applications to cranberry bogs (Winkler et al. 1996), and in neo tropical montane regions where amphib ians are declining, chloro thalonil has been regularly detected at levels that caused signifi cant mortality in the present study (Daly et al. 2007). This makes chloro thalonil a plausible contributor to declines, although additional work is needed to demon strate a causal link. Given these findings and similarities between the vital systems of amphibians and humans, we encourage future studies to quantify the effects of chloro thalonil on amphibian popu lations and human health.