• Amphibian;
  • FETAX;
  • Risk assessment;
  • Acute toxicity;
  • Comparative toxicity


  1. Top of page
  2. Abstract
  8. Acknowledgements

An expert workshop recently was convened to consider the frog embryo teratogenesis assay-Xenopus (FETAX) as a screening method for identifying the potential developmental toxicity of single chemicals and chemical mixtures. One recommendation from the workshop was that, in order to determine the utility of FETAX for ecological risk assessments, additional consideration of how the assay is conducted is necessary. In addition, a comparative evaluation would be useful of FETAX endpoints (i.e., survival, malformations, growth) versus each other, endpoints from aquatic toxicity tests using more commonly tested species of cladocerans and fish, and tests with other amphibian species. This review provides an evaluation and critique of the current FETAX protocol from two perspectives: Practical considerations relative to conducting the test and sensitivity of the assay (and associated endpoints) compared to tests with other species. Several aspects of the current standard protocol, including test temperature, diet, loading rates, and chemical exposure options, need to be modified to ensure that the assay is robust technically. Evaluation of FETAX data from the open literature indicates that growth is the most sensitive endpoint in the assay, followed by malformations and then survival; unfortunately, the growth endpoint often is not considered or reported in the assay. Comparison of FETAX data with acute toxicity data from tests with other amphibians or traditional aquatic test species indicates FETAX is relatively insensitive. This suggests that environmental risk assessments using acute hazard data from tests with traditional aquatic test species usually would be more protective of native amphibian species than risk assessments that use hazard data from FETAX.


  1. Top of page
  2. Abstract
  8. Acknowledgements

Dumont et al. [1] described a basic protocol for a toxicity test with the amphibian Xenopus laevis, the South African clawed frog. The test, commonly referred to as frog embryo teratogenesis assay-Xenopus (FETAX), was designed to serve as an alternative to mammalian models for detecting developmental toxicants, including teratogens. Xenopus laevis is an attractive species for this purpose because it is cultured and handled easily in a laboratory setting, and a relatively strong developmental biology knowledge base exists [2,3]. Standard methods have been described for FETAX [4,5], and comparative interlaboratory testing has been conducted using the standard protocol [6–10]. The test is initiated with blastula-stage embryos exposed to test solutions, which typically are renewed every 24 h. To help detect chemicals whose toxicity is enhanced by oxidative metabolism, an approach has been proposed in which test solutions can be modified with a metabolic-activation system derived from rats [11,12]. Mortality is assessed daily through the end of 96 h (completion of organogenesis) and surviving embryos are examined microscopically for overt malformations and growth (length) inhibition. Data are assessed via conventional techniques, such as calculation of median lethal concentration (LC50) or median effective concentration (EC50) values, or through comparison (usually as a ratio) of the test chemical concentration causing lethality versus the concentration resulting in malformations (teratogenic index). An additional endpoint that may be reported is the minimum (test chemical) concentration to inhibit growth (MCIG).

Although much of the focus on FETAX development and application has been as a screen for chemicals that might cause adverse developmental effects in mammals, the test also has been applied to evaluation of the potential ecological effects of single chemicals, defined chemical mixtures, and unknown toxicants in environmental samples [13–19]. Based on these types of studies, it has been proposed that FETAX should be used routinely in ecological risk assessments (ERAs) to augment currently used assays with other freshwater aquatic species [16,20,21]. The plausibility of this proposal was one topic of discussion at a 2001 workshop on FETAX, at which it was concluded that a comprehensive evaluation of the assay application to environmental assessments was lacking ( The purpose of this paper, therefore, is to provide a critical evaluation of the application of FETAX to ERAs.


  1. Top of page
  2. Abstract
  8. Acknowledgements

FETAX in ERAs: Practical considerations

Routine use of any bioassay in a regulatory setting necessitates optimization and standardization of the protocol in the context of its intended use. In the case of FETAX, this dictates consideration of factors such as organism quality (source, culture methods, diet, etc.), test conditions (temperature, test volume, organism loading rate, etc.), procedural aspects of chemical exposure (static-renewal vs. flow-through, measured vs. nominal chemical concentrations, documentation of metabolic activation system [MAS] effectiveness, etc.) and data collection/interpretation. Because the FETAX assay theoretically could be useful both for prospective and diagnostic (or retrospective) assessments, these factors must be addressed relative to testing chemicals both in well-defined media and complex environmental samples (often a poorly defined matrix). To address this, in the first part of this review, several practical aspects of FETAX are discussed that, for various reasons, may affect utility of the assay in ERAs.

FETAX in ERAs: Conceptual considerations

The primary justification for inclusion of an assay in a suite of biological tests for ERAs is that it provides unique information relative to the assessment or measurement endpoints of interest. One of the most common arguments for use of FETAX in ecological assessments is that X. laevis, as evaluated in the assay, represents amphibians, a class of organisms that currently appears to be under considerable environmental stress. Specifically, there is increasing evidence of declines and, in some cases, extinction of amphibian species throughout the world [22–24]. A variety of stressors, including habitat alteration, disease, and anthropogenic contaminants, have been proposed as potentially responsible for these declines [22,25–31]. Another observation that has focused attention on amphibians as currently at risk has been a seemingly increased incidence of hind limb malformations in a number of Rana species in North America. Although a variety of stressors may contribute to these malformations, the possible role of toxic contaminants has received the most attention [19,32–35]. The possible contribution of environmental contaminants to adverse effects in amphibians, such as population declines or malformations, is difficult to assess, in part, because of a lack of standardized toxicity tests with species in this class, and resultant limitations in high-quality toxicology data suitable for either prospective or diagnostic assessments [36,37]. Because many amphibian species of concern (e.g., North American ranids) are difficult, if not impossible, to work with in the laboratory, the use of X. laevis as a broad testing surrogate for amphibian species has received much attention. However, a critical uncertainty in the use of FETAX as a representative model of amphibians is that little is known concerning the relative sensitivity of this assay to toxicants, either from a qualitative or quantitative perspective, compared with other species at potential risk or relative to more traditional aquatic test species. Although there has been a substantial amount of developmental biology research with X. laevis, there are few comparative toxicology data for this species relative to other amphibians.

Another argument for routine use of FETAX in ERAs involves the fact that the test focuses on early embryonic development. This can be a uniquely sensitive period with respect to sensitivity to contaminants. This fact is exploited in commonly used embryo-larval tests with various fish species [38–42]. What these standard fish assays typically do not include, relative to the FETAX protocol, is formal examination of test organisms for morphological abnormalities/malformations at test completion. Most fish tests rely on nonlethal developmental effects to be expressed as alterations in growth or external morphological abnormalities. Hence, FETAX potentially could be useful in providing insights into the occurrence of developmental effects if such effects could be linked to specific developmental toxicants with specific modes of action.

An important determinant of the utility of FETAX in ERAs is whether those unique aspects of the assay (i.e., an amphibian model of early development) result in it being more sensitive than other commonly used species/tests. If FETAX consistently is less sensitive than other test systems, its use in a suite of assays might not be justified based on animal welfare, scientific, and cost consideration. Further, if FETAX is relatively insensitive, it would be imprudent to replace existing aquatic toxicity assays with this test. However, there has not been a systematic evaluation of the sensitivity of FETAX compared with other test methods using aquatic organisms. As part of a critical evaluation of the utility of FETAX, we present results of an analysis in which the relative sensitivity of FETAX endpoints are compared internally (i.e., 96-h LC50 vs. 96-h EC50 malformations; 96-h EC50 malformations vs. 96-h MCIG). We also present comparisons for single chemical tests with typical aquatic test species used to develop U.S. Environmental Protection Agency ambient water quality criteria (AWQC), comparisons for single chemical tests conducted with other species of amphibians, and comparisons for a limited number of tests of complex environmental samples conducted with a variety of aquatic test species.


  1. Top of page
  2. Abstract
  8. Acknowledgements

Practical considerations

In this section, we consider several factors that could affect results obtained using the currently recommended FETAX protocol [4,5]. Not every test parameter that could be considered is addressed; rather, we have attempted to focus on those that potentially could be most important in affecting data quality, accuracy, and test interpretation.

Organism quality

To ensure generation of accurate, reproducible data, consistent organism quality is of paramount importance. The best approach to producing known quality test animals is through utilization of standard, well-documented culture conditions; ideally, this would be complemented by the use of a suite of performance-based criteria (e.g., growth, reproductive output, embryo viability, sensitivity to reference toxicants) to ensure that the culture conditions and resultant test animals consistently are robust. Many examples can be found of this approach for other fish and invertebrate species used for regulatory testing in the United States [39,40,43]. Although X. laevis has been used for developmental biology research for many years, there are no standard approaches for culturing this species. For instance, basic water quality conditions and diet vary greatly among laboratories culturing and/or testing Xenopus. As an example, in recent personal communications with five different groups that culture X. laevis for commercial purposes, one of the authors of this paper (G. Ankley) found a large variation in the diets used to maintain the adults, including salmon or trout starter from three different suppliers, a commercially available frog food, a specially formulated frog food, and/or fresh meat products such as beef liver. A published guide for rearing X. laevis recommends use of ground beef liver or lung to maintain adults [4,5]. The effects of diet alone on endpoints such as growth is of importance, particularly for interlaboratory comparisons, given the observation that different diets may produce a greater than 10-fold difference in larval dry weight for X. laevis [44]. Part of the reason for the large variation in culture conditions, as exemplified by these different diets, undoubtedly involves a lack of resources to support the basic biological research necessary to optimize/standardize rearing methods for this species. However, if X. laevis is to be used routinely for regulatory testing, either for FETAX or some other protocol, research of this type must be a high priority. Other researchers also have discussed the lack of scientific data on the housing and culture conditions necessary to optimize the health and fecundity of this species in captivity [45].

Unlike most commonly used mammalian species, genetic composition of organisms used for aquatic testing can vary greatly among different stocks of animals, potentially resulting in significant interlaboratory variation in observed sensitivity of a given species to test chemicals [46–49]. Genetic lineage is confounded, at times, by record keeping that is inadequate to trace specific lineages of original brood stock. Xenopus laevis does not escape this shortcoming. Further, because X. laevis is a tetraploid, there has been speculation that this could contribute to greater divergence in genetic characteristics among isolated populations than might occur in diploid species ( Although a number of X. laevis cultures currently are maintained in the United States, little is known about the genetic characteristics or lineage of many of these animals, and it has not been a routine practice for laboratories to exchange animals for the explicit purpose of out-breeding to maintain genetic diversity. This is another area that will require attention if FETAX is to achieve routine use in ERAs, particularly those driven by specific regulatory mandates.

Test conditions

To ensure the generation of valid data, recommended conditions for any biological test method must be broadly based on knowledge of optimal requirements of the test organism. However, within the context of this requirement, there may be significant leeway in developing standard test conditions that can achieve desired results. For example, a species may be tolerant of a comparatively wide range of temperature such that any of a number of discrete temperatures could be selected for a standard test. An important requirement associated with this selection, however, is demonstration and documentation that the conditions selected do not have the potential to affect test performance through confounding effects on the biology of the species of concern. Table 1 summarizes test conditions for the standard FETAX protocol recommended by the American Society for Testing and Materials [4,5]. The majority of recommended conditions would appear to be reasonable with respect to X. laevis biology; however, three variables are of potential concern relative to test validity/interpretation. These variables are test temperature, test-organism loading, and, in the case of poorly characterized environmental samples, identification of acceptable ranges of water-quality characteristics.

The FETAX is conducted at 24 ± 2°C (Table 1); however, it has been recommended that breeding adults used to generate test embryos be cultured at 21 ± 2°C [5]. This could result in a potential temperature differential between the culture and test conditions of as much as 7°C. We are not aware of any other routinely used short-term toxicity test with fish or aquatic invertebrates that recommends (or would allow) this type of differential between culture and test temperatures. The ability of poikilothermic organisms to survive and adapt to sometimes widely varying environmental temperatures is well-established [50,51]; however, the biochemical and physiological mechanisms underlying temperature compensation usually require a period of acclimation that would not be afforded in conjunction with collecting newly fertilized embryos and conducting FETAX. An additional concern involves the allowable temperature range during FETAX; conducting the test in the range of 22 to 26°C helps ensure that the test animals will complete organogenesis (stage 46) [52] within 96 h [4,5]. However, temperatures above 26°C can cause an elevated rate of malformations in control embryos [4,5]. It does not seem technically defensible (or logistically reasonable) to have the upper limit of the recommended assay temperature correspond directly to a temperature that could result in an unacceptable malformation rate. In addition, a temperature range of 5°C (22–26°C) as permitted in the existing FETAX method [5] is likely to introduce additional variability into test endpoints. This temperature range also is greater than is permitted in many other aquatic test methods (e.g., [39,40]). Given the uncertainty introduced by the differential between culture and test temperature, the upper range of acceptable temperature during FETAX, and the potential variability introduced by temperature variation, further research concerning this test parameter would be advisable.

Table Table 1.. Summary of recommended test conditions for frog embryo teratogenesis assay-Xenopus
Test typeStatic-renewal (daily)
Water temp.24 ± 2°C
Photoperiod12:12-h light:dark
Test solution volume10 ml
Dilution/control waterFETAX solution
Solution pH7.7 (range, 6.5–9.0)
Organisms stageMid-blastula (stage 8) to early gastrula (stage 11)
No. of test concentrationsMinimum of five
No. of replicates/treatmentTwo/treatment; four/control(s)
Reference toxicant6-aminonicotinamide
Metabolic activation (optional)Hepatic microsomes from Aroclor 1254-treated rats
Test duration96 h
Test endpointsMortality, malformations, growth length inhibition
Test acceptability>90 survival, <10% malformations in control(s); >90% of control organisms as stage 46 at test conclusion

Test-organism loading (i.e., organism mass-to-test solution ratio) probably is the single most critical variable affecting water quality in aquatic bioassays. Water quality parameters that can affect organism health (i.e., bias test results) and are most influenced by test-organism loading are dissolved oxygen content and concentrations of potentially toxic metabolic byproducts, principally ammonia. In recognition of this, relatively specific guidelines have been developed for test-organism loading in both static and flow-through tests with fish [53]. The recommended test-organism loading for FETAX is 25 embryos per 10 ml of solution (Table 1). Assuming a wet weight of 10 mg/embryo, this translates into a test-organism loading of approximately 25 g/L, substantially above the 0.5 to 1 g/L recommended for static-renewal bioassays at 24°C with fish or other amphibians [53]. The advantage of small volumes of test solution for FETAX (or any assay) primarily is one of logistics related to the generation of toxicant stocks and their subsequent disposal. However, this is not an advantage if water quality is compromised. Previous studies have evaluated the effect of test solution volume on the results of assays with X. laevis. Schuytema et al. [54] exposed X. laevis embryos to azinphos methyl and a formulation of azinphos methyl using the recommended test-organism loading for FETAX (25 embryos/10 ml), as well as loadings of 10 and 20 embryos/100 ml. Embryos exposed to a range of azinphos methyl concentrations in 10-ml test solution volumes exhibited greater mortality, higher incidence of deformities, and decreased size compared to embryos exposed in 100-ml test solution volumes. Similar observations were made in a study of the effects of experimental design on evaluation of contaminated sediment and soils [55]. In that study, FETAX conducted at a test-organism loading of 20 embryos/20 ml of aqueous sediment or soil extract exhibited equivalent or greater developmental effects than when tested at a loading of 20 embryos/35 g of sediment/soil in 140 ml of FETAX solution (moderately hard reconstituted water) [56].

Recent studies have indicated that, under standard FETAX conditions, concentrations of total ammonia as great as 6 mg/L can occur between daily renewals, with amounts tending to increase toward the end of the 96-h assay as the X. laevis metabolic rate increases [57]. At the pH normally used for FETAX, these concentrations are below those that should cause lethality or gross malformations in FETAX during a 96-h assay [57,58]. However, observed concentrations are approximately five times higher than the chronic ammonia water-quality criterion for aquatic life at a comparable pH [59]. The possibility of sublethal effects (e.g., growth) other than gross malformations associated with this amount of ammonia accumulation, and/or adverse effects in solutions at higher pH values, have not been assessed in the FETAX protocol. Further, the potential exists for ammonia to exert toxicity in an interactive fashion with test chemicals of concern. Hence, it would seem reasonable, based solely on water-quality considerations, that the test-organism loading recommended for FETAX be reduced. In addition, given the potential for previous FETAX results to be biased in a conservative manner by the undocumented accumulation of ammonia, caution should be used with respect to quantitative interpretation of historical results from the assay.

Dumont et al. [1] originally envisioned FETAX as suitable for assessing developmental toxicity of chemicals under controlled conditions, including a well-characterized water-quality matrix. To support this type of testing, a moderately hard reconstituted water known as FETAX solution was developed [56] and currently is recommended for the standard protocol (Table 1). However, FETAX increasingly has been used in diagnostic ERAs with samples such as effluents, surface and well waters, and aqueous sediment extracts [15,17–19]. Although this may be a logical extension of the assay, water samples from the environment often can be unusual with respect to basic water-quality parameters. If these parameters are not measured, and/or their effects on biology of developing X. laevis are not known, the potential exists for biased and/or misinterpreted test results. For example, in a recent study, Tietge et al. [57] assessed possible causes of adverse effects in samples from a site reported to cause developmental toxicity in FETAX [18,19]. Water from the site caused developmental delays, as well as what appeared to be cranio-facial and gut abnormalities in the larval X. laevis [18,19,57]. However, these effects appeared not to be caused by toxic chemicals but by ion (possibly potassium) deficiencies in the rather soft surface water [57]. This type of ion deficiency appears to have caused positive (toxic) results in FETAX with samples not just from this site, but others as well [60]. In controlled studies, Luo et al. [61] investigated the effects of Mg+2 on teratogenic effects of various metals using FETAX. Although the standard FETAX solution (which contains Mg+2 at a concentration of 620 μmol/L) produced a control malformation rate of 5.4%, decreasing the Mg+2 concentration to 62, 6.2, and 0 μmol/L produced control malformation rates of 32, 100, and 100%, respectively. Less than 5% mortality was observed in all but the 0% Mg+2 control treatment where a greater mortality rate was noted. Based on these observations, it is clear that additional research is needed to better characterize the effects of variations in common water quality parameters (e.g., ion composition, pH, dissolved and total organic matter, etc.) on FETAX results. This information, in conjunction with routine characterization of key water-quality parameters in environmental samples before and during the assay, would strengthen the utility of FETAX for diagnostic ERAs.

Aspects of chemical exposure

The FETAX can be used to generate data for single chemicals and/or defined mixtures to support prospective ERAs [20]. The current protocol recommends a static-renewal regime in which stocks of chemicals are renewed on a periodic (generally 24-h) basis (Table 1). An important problem and test design deficiency is that the majority of FETAX assays conducted to date have not routinely measured typical water-quality parameters (hardness, dissolved oxygen, ammonia, etc.) or chemical concentrations to which the developing embryos are exposed. The use of a static-renewal regime without supporting analytical chemistry to confirm exposure has the potential to introduce significant uncertainties relative to interpretation of assay results. Specifically, many chemicals of concern, including those prone to degradation in aqueous solution, relatively hydrophobic compounds (e.g., those with a log10 octanol-water partition coefficient (Kow) greater than 3.5), and volatile chemicals may be poorly suited to static-renewal testing. When testing these types of chemicals, a static-renewal regime results in a series of pulsed exposures, with nominal (target) concentrations sometimes achieved only for minutes or hours during the assay [35]. The net result of this is that, without a significant number of chemical measurements (perhaps more than once daily), results from static-renewal assays with unstable, hydrophobic, and/or volatile chemicals will associate adverse effects with higher chemical concentrations than those actually experienced by the test organisms. This, of course, would underestimate the potential ecological impacts of chemicals of concern based on measurement or prediction of their concentration in aquatic environments. To avoid this tendency toward type II error (i.e., the potential to underpredict impacts), virtually all current testing for the purpose of quantitative prediction of risk utilizes flow-through assays with constant chemical delivery and/or requires concurrent analytical verification of chemical exposure ([41,42,62–65]; OPPTS_ Harmonized/850_Ecological_Effects_Test_Guidelines/index. html). Based on current practice, it would seem inadvisable to use historical data generated using FETAX for quantitative risk assessments unless chemical exposure concentrations were documented using appropriate analytical procedures. Further, if the basic FETAX protocol is used in future testing, it must be modified to enable routine use of a flow-through exposure regime for chemicals whose properties might result in fluctuating exposures, or sufficient chemical measurements must be conducted to characterize adequately test-organism exposure to the chemical of concern over the duration of the test.

Due to logistical constraints, diagnostic assessment of the toxicity of samples from the environment, including effluents, surface water, and sediments (or sediment extracts), often necessitates use of static-renewal as opposed to flow-through exposures. For example, in the United States, regulatory testing of complex effluents almost solely is comprised of static-renewal assays with fish and invertebrates [39]. This limits the ability of these tests to detect effects associated with the types of chemicals mentioned above, i.e., those that degrade rapidly, are hydrophobic, and/or are volatile. Other than more frequent renewals, little can be done to better optimize detection of rapidly degrading compounds. In some cases, it may be possible to test volatile compounds using a zero headspace test design, while the ability to detect effects associated with hydrophobic chemicals can be enhanced through the use of larger test solution volumes.

It may be possible to compensate for the effects of losses in static-renewal tests by using larger sample volumes. Specifically, disappearance of a hydrophobic test chemical from an aqueous solution can be predicted based upon its Kow as a log linear function of the total mass of chemical relative to the lipid biomass of organisms in the test chamber, assuming the test organisms are the main partitioning phase. In the recommended FETAX protocol, in which 25 organisms (lipid biomass of 4 mg) are held in 10 ml of solution, only about 20% of a chemical with a log Kow of 4.0 would be present at the end of a 24-h exposure period (D. Mount, U.S. Environmental Protection Agency, Duluth, MN, personal communication). This effect, of course, becomes more exacerbated for chemicals with higher log Kow values. However, merely by increasing the volume of the test solution (i.e., the total mass of chemical available), it is possible to prolong significantly exposure to the desired concentration of the chemical of concern. For example, about 71% of the chemical with a log Kow of 4.0 would be predicted to be present at the end of 24 h when 25 embryos are held in 100 ml instead of 10 ml of test solution. This observation, in conjunction with biological issues associated with test volume [57], reinforces the need to modify biomass to solution ratios for the FETAX.

A criticism of FETAX as a surrogate model for identifying mammalian teratogens and developmental toxicants has been that the developing frogs possess only a limited capacity to metabolize test chemicals to more active forms. To address this, a MAS was developed using hepatic preparations from rats treated with chemicals, such as Aroclor 1254, to induce cytochromes P450 responsible for the oxidative metabolism of xenobiotics [12]. The assay then can be conducted with the chemical(s) of concern, with supplemental addition of MAS, a source of NADPH, and antibiotics. Based on changes in toxicity (mortality and malformation rates), there is some indication that the MAS system is effective in metabolic activation of relatively reactive chemicals, such as cyclophosphamide and benzo[a]pyrene [12,65]. However, there has been little analytical verification of the effectiveness of the MAS in terms of actual metabolites produced or maintenance of any given level of activity over time. Given the instability of membrane-bound cytochromes P450 [66], it is questionable whether a significant level of oxidative activity would be maintained for more than a few minutes in a physiologically dilute medium that is about 12°C cooler than the body temperature of the animal from which the enzymes were derived. This leads to uncertainty in terms of exposure-response relationships derived using FETAX, from both qualitative (what) and quantitative (how much) perspectives. One way to address this uncertainty would be to perform kinetic studies with FETAX using a suite of chemicals that undergo biotransformation to known products. In addition, this uncertainty could be assessed through better analytical verification of either disappearance of the parent compound or production of metabolites in the test system. It may be that use of the MAS in FETAX for addressing ERA issues is less important than when the assay is used as a screen for developmental toxicants in mammals [5]. However, if the MAS approach is to be used as part of a standard protocol for testing chemicals, further attention needs to be given to characterizing and standardizing this aspect of the assay.

Data collection

The primary endpoints in FETAX are determinations of survival concomitant with the daily renewal of the test solution and assessment of malformations in surviving animals at test conclusion. Relative growth (embryo length) also may be determined at termination of the assay for calculation of the MCIG value. From the standpoint of mechanism/mode of action, the data potentially most unique to the FETAX assay (relative to other more commonly used assays with aquatic species) are malformation incidence and type(s). Given this, there are approaches whereby the quality of malformation data collected in conjunction with the assay could be improved. Specifically, although there is a well-developed description of common gross malformations in 96-h-old X. laevis [67], a more exhaustive evaluation of morphological abnormalities could be achieved in order to extend the sensitivity of the assay through detection of more subtle, but meaningful, malformations that might occur at lower chemical concentrations. Increasing the frequency and intensity of morphological evaluations would have consequences in terms of the expense associated with and the expertise required to perform FETAX. However, if the assay has the potential to bring unique toxi-cological information to ERAs, these consequences may be justified.

Another approach to enhance utility of the malformation endpoint would be to assess morphological abnormalities in animals that die (or clearly will die) during the assay. The present FETAX protocol does not recommend explicitly evaluation of organisms scored as dead during the assay. However, to be useful as an assay endpoint, particularly for ecological risk assessment, the malformation endpoint of FETAX must provide either or both of the following: Information on developmental effects not identified through other FETAX endpoints or other more commonly conducted aquatic toxicity tests, and information providing causality of observed effects (i.e., malformation syndromes that can be linked to specific chemicals, classes of chemicals, or modes of action).

One aspect of our current analysis was assessment of the sensitivity of the FETAX 96-h LC50 endpoint versus the 96-h malformation EC50 and the 96-h MCIG. We reviewed numerous studies to compile data for these endpoints from FETAX studies of pure chemicals (references provided upon request). Although the 96-h LC50 and malformation EC50 typically are reported for most FETAX studies, the 96-h MCIG is reported less frequently. Figure 1 presents the comparison of the 96-h LC50 versus the 96-h malformation EC50 (n = 90) and, based on the shifting of the points to the right of the hypothetical 1:1 correspondence line, clearly supports the hypothesis that the 96-h malformation EC50 is more sensitive than the 96-h LC50. When the malformation EC50 is compared with the 96-h MCIG (Fig. 2), it is apparent that the malformation EC50 does not provide increased sensitivity relative to the growth endpoint. In fact, it appears that, in the majority of the total comparisons (n = 68), the growth endpoint actually is slightly more sensitive than the malformation endpoint. Other investigators also have noted that growth effects in FETAX occur at chemical concentrations that are lower than those required to produce malformations [16,56]. The potential ramifications of these observations are important given the increased level of labor, time, and cost necessary for collection of malformation data relative to a simple measurement of growth (i.e., length).

thumbnail image

Figure Fig. 1.. Frog embryo teratogenesis assay-Xenopus (FETAX) 96-h median lethal concentrations (LC50) versus FETAX 96-h median effective concentrations (EC50) for developmental effects (n = 90).

Download figure to PowerPoint

Several additional noteworthy observations are relative to the malformation endpoint in FETAX. Specifically, while reviewing the extant literature concerning FETAX, we were struck by the wide variation in malformation rates reported for control embryos. Control malformation rates in studies that we reviewed ranged from 0 to 50% [68] with frequent reports of rates between 5 and 10%. Not only does this seem like a relatively large range of malformations for control organisms, but the occurrence of malformations in excess of environmental background rates for amphibians of <1 to 2% [69,70] might indicate an unidentified problem(s) with the assay. The acceptable malformation rate of up to 10% in the FETAX protocol [5] itself appears to be high for a developmental assay. In any event, it also should be mandatory for control malformation rates to be reported explicitly from all studies. Many studies either do not report these data [71] or report that test data somehow were corrected for control malformation rate without reporting this value [72,73]. Further consideration of possible causes of large and/or variable malformation rates in controls from FETAX is warranted.

thumbnail image

Figure Fig. 2.. Frog embryo teratogenesis assay-Xenopus (FETAX) 96-h median effective concentrations (EC50) for developmental effects versus FETAX 96-h minimum concentrations inhibiting growth (MCIG) (n = 68).

Download figure to PowerPoint

Comparative sensitivity

In this section, we assess results from FETAX relative to acute data from tests with other amphibians, as well as aquatic species (primarily cladocerans, fish) used more commonly for ERAs. Variations exist among the assays, with respect to how they were conducted, that might affect their observed sensitivity apart from variations in species/life-stages/endpoints. For example, many of the non-FETAX tests were conducted using flow-through systems and measured chemical concentrations that, even if there were comparable sensitivity among species, likely would result in effects concentrations lower than those using the recommended (static-renewal, unmeasured) FETAX protocol. On the other hand, unacceptable water-quality excursions (e.g., elevated ammonia concentrations) might serve to increase observed sensitivity of FETAX relative to other assays, again in the absence of any inherent differences in actual species sensitivity to the test chemicals of interest.

With these caveats in mind, there are several possible comparisons that can be made to evaluate the utility, and comparative sensitivity, of data generated using FETAX. These include comparison of FETAX sensitivity to single chemicals with results of early life-stage tests conducted with other amphibians exposed to the same chemicals; comparison of FETAX data for single chemicals with results obtained from tests with cladocerans and fish commonly used for water quality criteria testing; and comparison of FETAX results with those for other species in exposures involving complex environmental samples.

When comparing single chemical acute toxicity data from FETAX versus other amphibian species, it became evident that there are only limited data of similar nature available for this assessment. Tests with X. laevis have been conducted with a greater emphasis on the use of FETAX as a screen and surrogate for mammalian tests in assessing developmental effects. As a result, a large portion of the FETAX data are comprised of tests on compounds of potential interest as mammalian developmental toxicants. Conversely, the limited amount of data available for other amphibians are from tests with chemicals that primarily are of interest as environmental contaminants. Hence, synoptic data for these types of comparisons are limited; however, Schuytema and Nebeker [74] compiled an extensive toxicity database for the development of AWQC for amphibians. Unless otherwise noted, the data used in this manuscript for comparative purposes were gathered from the pertinent references listed by Schuytema and Nebeker [74]. Figure 3a through e presents a comparison of FETAX 96-h LC50 values for aluminum, chloroform (9-d LC50), mercury, dieldrin, and azinphos methyl versus those for other amphibian species for which similar data were available. The aluminum data are from studies started with embryos, and data for the other compounds are from tests initiated with tadpoles of a similar age. These data indicate that, for all comparisons, tests with native amphibian species appear to be more sensitive than FETAX. Although this comparison is somewhat constrained by the limited data available from 96-h tests during early development, other investigators have reported similar results from various types of study designs [14,75,76]. Perhaps the largest such data set in existence are the early life-stage toxicity data generated by Birge et al. [77] to classify the sensitivity of 25 amphibian species, relative to the rainbow trout (Oncorhynchus mykiss), to a series of 34 metals and inorganic elements and 27 organic chemicals. In these experiments, embryos were exposed to test chemicals from fertilization through 4-d posthatch; as a result, the test duration was species-specific due to varying species-specific developmental rates. Based on their experimental design, X. laevis was the most tolerant of the eight amphibian species evaluated for their sensitivity to organic chemicals.

thumbnail image

Figure Fig. 3.. Comparative 96-h median lethal concentrations (LC50) for amphibian embryos exposed to (a) aluminum, (b) inorganic mercury, (c) dieldrin, and (d) Guthion, and (e) 216-h LC50 values for chloroform.

Download figure to PowerPoint

A second comparison made to help assess the utility of FETAX for use in ERA was evaluation of single chemical test results for FETAX versus traditional aquatic test species, in this case, those used to develop AWQC published by the U.S. Environmental Protection Agency. Data for species other than X. laevis (i.e., FETAX) were taken from the chemical-specific AWQC documents [59,78–89] or other references as indicated in Figures 4, 5 to 6. These AWQC documents contain the references to the primary data sources. The comparisons were conducted using the species mean acute values from the AWQC documents and the 96-h LC50 (and 96-h EC50 and MCIG, if available) from FETAX studies. Figure 4a through g contains the results of comparisons for aluminum, cadmium, copper, nickel, selenium, mercury, and zinc. Figure 5a through c contains the results of the comparisons for three industrial chemicals, ammonia, aniline, and pentachlorophenol. Figure 6a through c contains the results for comparisons of the three pesticides atrazine, malathion, and parathion.

thumbnail image

Figure Fig. 4.. Comparative frog embryo teratogenesis assay-Xenopus (FETAX) 96-h median lethal concentrations (LC50) or 96-h minimum concentrations inhibiting growth versus U.S. Environmental Protection Agency ambient water quality criterion 96-h median lethal concentrations (LC50) or species mean acute values (SMAV) for (a) aluminum, (b) cadmium, (c) copper [17,21,99], (d) nickel, (e) selenium, (f) inorganic mercury, and (g) zinc [17,21].

Download figure to PowerPoint

Overall, FETAX was not the most sensitive test in any comparison based on LC50 values, with the exception of aluminum. When possible, the more sensitive FETAX MCIG values also were included in the comparisons and, in these instances, the growth endpoint was the most sensitive endpoint for nickel and pentachlorophenol. Although mortality in FETAX was more sensitive to aluminum than Ceriodaphnia dubia 48-h mortality (the most sensitive freshwater species from the AWQC dataset), FETAX still was the least sensitive of the amphibian tests conducted with aluminum (Fig. 3). Mortality in FETAX conducted with nickel was slightly less sensitive than that of Daphnia magna (48-h EC50 value), but the FETAX MCIG was approximately 10-fold lower than the acute D. magna LC50. In all other comparisons of metal toxicity in FETAX versus the AWQC species, a crustacean was ranked as either the first or second most sensitive species, with the exception of cadmium. A salmonid fish was most sensitive to cadmium, and D. magna was the second most sensitive species. Although inclusion of MCIG values seemed to increase the sensitivity of FETAX relative to other assays/species, it is important to note that inclusion of a comparable short-term growth endpoint for the AWQC organisms likely would increase proportionally the sensitivity of those tests as well.

For toxicity comparisons using the industrial chemicals ammonia, aniline, and pentachlorophenol, fish species were the most acutely sensitive to ammonia and pentachlorophenol, and the cladocerans D. magna and C. dubia were most sensitive to aniline. Inclusion of the FETAX MCIG in the comparison resulted in this endpoint being more sensitive than the fish LC50 for pentachlorophenol. For the three pesticides, mortality of the midge Chironomus tentans was the most sensitive endpoint for atrazine, and the bluegill sunfish and D. magna were most sensitive to malathion and parathion, respectively. Therefore, based on the comparisons presented here, it appears that traditional aquatic tests species generally are more sensitive than FETAX when comparisons are based on 96-h lethality data.

A final evaluation of the utility of FETAX for use in ERAs involves comparison of results of toxicity tests with complex environmental samples. Even fewer comparative data exist for synoptic studies of environmental samples using FETAX and other aquatic test species than for single chemical studies. In one example, Dawson et al. [15] used FETAX and short-term fathead minnow (Pimephales promelas) assays to assess metal-contaminated sediments. The fathead minnow typically was most sensitive, with fish EC50 values for malformations and growth four- to sixfold and 1.5- to ninefold lower than for FETAX. The lethality endpoint in fish was approximately 10-fold more sensitive than lethality in FETAX. Birge et al. [90] used FETAX and embryo-larval tests with rainbow trout, blue-gill sunfish (Lepomis macrochirus), and fathead minnow to evaluate the toxicity of two complex industrial effluents. Rainbow trout, fathead minnow, and bluegill sunfish were three-to 10-fold, 1- to 1.5-fold, and 0.5- to twofold more sensitive than FETAX in tests of the two effluents. In a more recent study, Fort et al. [91] also used FETAX and a short-term fathead minnow teratogenesis assay to assess waste site soil extracts. The fathead minnow assay always was more sensitive than FETAX in terms of embryo lethality. The fish was approximately 0.5- to fourfold more sensitive than FETAX in terms of malformations, and from approximately fivefold less sensitive to fourfold more sensitive when evaluating a growth endpoint.


  1. Top of page
  2. Abstract
  8. Acknowledgements

Routine use of FETAX has been proposed in ERAs to augment assays with other freshwater aquatic species [16,20,21]. This proposal, along with many other aspects of FETAX and more specialized X. laevis testing, was discussed at a focused workshop at which it was concluded that a comprehensive evaluation of the application of FETAX for ERAs was lacking ( As a foundation for such an evaluation, we critically evaluated test conditions recommended for FETAX and compiled data on the comparative sensitivity of FETAX relative to more traditional aquatic test species and other amphibian species. An important determinant of the utility of FETAX in ERAs is whether those unique aspects of the assay (i.e., an amphibian model of early development) result in it being more sensitive than other commonly used aquatic test species or other amphibian species.

In our review of FETAX, a number of uncertainties/shortcomings were identified relative to recommended test methods. For example, assay standardization is hampered due to the lack of uniform dietary requirements and well-characterized stocks of test organisms. The results of FETAX also are poorly understood with regard to the effects of several key water-quality parameters, including temperature and basic ion composition; the latter would appear to be particularly problematic when testing poorly characterized complex mixtures from the environment. The test-organism loading used for FETAX also presents serious problems, both from the standpoint of maintenance of adequate water quality (e.g., to avoid accumulation of waste products) and delivered chemical dose. This test organism loading issue could be addressed either through significant increases in test volume or, preferably, by conducting tests under flow-through conditions. In either case, FETAX would be conducted best with a focus on using measured, as opposed to nominal, test chemical concentrations as a basis for reporting results.

Endpoints generated from FETAX may need to be reconsidered. Based on our analysis of existing data, a growth endpoint, such as the MCIG, should be assessed and reported routinely, although this frequently is not the case in existing studies. Our evaluation indicates that the FETAX 96-h EC50 for malformations is more sensitive than the 96-h LC50. This was not unexpected, because various nonlethal malformations occur at concentrations less than those killing 50% of the test organisms. More surprising from our analysis was that, when the 96-h malformation EC50 was compared with the 96-h MCIG, the MCIG appeared to be equally or more sensitive than the 96-h malformation endpoint. Other authors have reported that the FETAX MCIG is a more sensitive endpoint than the LC50 or EC50 for malformations. However, to our knowledge, this study presents the first comprehensive, systematic data evaluation that supports this general observation from individual studies. The FETAX MCIG endpoint has been underutilized and, in many studies, has not been reported because of the emphasis placed on the malformation endpoint.

thumbnail image

Figure Fig. 5.. Frog embryo teratogenesis assay-Xenopus (FETAX) 96-h median lethal concentrations (LC50) or 96-h minimum concentrations inhibiting growth (MCIG) versus U.S. Environmental Protection Agency ambient water quality criterion 96-h median lethal concentrations (LC50) or species mean acute values (SMAV) for (a) ammonia, (b) aniline, and (c) pentachlorophenol [99].

Download figure to PowerPoint

In considering the malformation endpoint specifically, it is clear that a better understanding of FETAX test conditions that might affect this response, relative to normal background rates in X. laevis, is critical if malformations are to be used as a technically defensible endpoint. For example, poorly controlled (elevated) test temperatures could contribute to elevated malformation rates, as could the accumulation of ammonia associated with excessive loading rates. Given these observations, there may be a greater disparity in comparative sensitivity between the actual malformation EC50 and the MCIG than indicated in Figure 2 (although it also is possible that the accumulation of waste byproducts could confound and artificially increase the sensitivity of the MCIG as well). Based on these observations, we feel it is quite possible that malformation and MCIG data from historical FETAX studies could overstate actual effects.

Comparisons of the results of FETAX with the limited acute toxicity test data available from similar studies with other amphibian species suggest that FETAX is an insensitive test for predicting the acute hazard of environmental contaminants to native North American amphibian species. In the limited number of comparisons possible based on available acute toxicity data, tests with other amphibian species always were more sensitive than FETAX. Similarly, X. laevis was the most tolerant of 25 amphibian species evaluated in a comprehensive suite of early life-stage studies of 34 metals and inorganic elements and 27 organic chemicals [77]. Therefore, short-term tests with X. laevis (e.g., FETAX) do not appear to be good surrogates for assessing the hazard and/or quantitative risk of chemicals to other species of amphibians.

Comparison of FETAX mortality and malformation data with the large AWQC acute toxicity data set for invertebrates and fish indicates that FETAX also is less sensitive than these organisms to environmental contaminants, with the exception of aluminum. Although FETAX endpoints were more sensitive to aluminum than acute tests with other aquatic species, results from FETAX indicated it was less sensitive than tests of aluminum with other amphibians. Amphibians, in general, appear to be more sensitive to the lethal effects of aluminum than other aquatic species; the mechanism(s) through which this differential sensitivity occurs is uncertain.

thumbnail image

Figure Fig. 6.. Frog embryo teratogenesis assay-Xenopus (FETAX) 96-h median lethal concentrations (LC50) or 96-h minimum concentrations inhibiting growth (MCIG) versus U.S. Environmental Protection Agency ambient water quality criterion 96-h median lethal concentrations (LC50) or species mean acute values (SMAV) for (a) atrazine [71], (b) malathion, and (c) parathion.

Download figure to PowerPoint

The data presented herein suggest that the utility of FETAX in ERAs is limited due to the usually greater sensitivity both of other amphibian species/assays and more commonly used aquatic test species such as invertebrates (e.g., cladocerans) and fish. However, if FETAX is not used as a surrogate species for native amphibians, there may be a perceived gap in existing risk-assessment tools given that amphibians can be receptor species of interest. Recent data suggesting both worldwide declines in various amphibian species and increased incidence of deformities in North American amphibians have focused considerable attention on this group of vertebrates. One of the implicit assumptions of this increased attention is that existing criteria developed to be protective of most invertebrates and fish (i.e., EPA AWQC are designed to be protective of 95% of species, 95% of the time) are not protective of amphibians. Limited efforts have been made to assess this assumption. For instance, in the amphibian and fish early life-stage studies conducted by Birge et al. [77] previously discussed herein, the investigators found that amphibian species frequently were more sensitive to metals than rainbow trout and used this observation to support the need for evaluation of contaminant effects on amphibians. However, evaluation of the EPA AWQC dataset for metals indicates that invertebrates (typically crustaceans), and not fish such as the rainbow trout, typically are the most-sensitive species to metals.

Other studies have examined more directly the protectiveness of existing AWQC for amphibians. In an evaluation of the effects of dieldrin [92] and azinphos methyl [54], the authors concluded that existing AWQC were likely to be protective of amphibians. In another study, evaluation of the effects of ammonia on native ranid species also led the authors to conclude that criteria based on effects on fish would be protective for many anuran amphibian species [93]. Similarly, in an evaluation of the effects of 4-nonylphenol, carbaryl, copper, pentachlorophenol, and permethrin on the southern leopard frog (Rana sphenocephala), the authors concluded that, “values obtained from rainbow trout may be conservative for many chemicals and therefore protective of amphibians” [94]. Based on these observations and the data evaluated in the present study, existing tests with more traditional aquatic test species appear to be more appropriate for protecting native amphibian species than data generated from FETAX. Other authors also have noted that “the literature on environmental contamination is devoid of compelling evidence that adult and larval amphibians are more sensitive to chemicals than other land and aquatic vertebrates” [95]. The question for scientists involved in ERAs is how best to assess the full range of potential stressors, including climatic changes and habitat destruction, to provide for adequate levels of protection for the largest number of species, including amphibians.


  1. Top of page
  2. Abstract
  8. Acknowledgements

In considering technical weaknesses of the existing FETAX protocol, as well as the relative insensitivity of results from FETAX compared to those generated with other aquatic species (including other amphibians), it is difficult to envision how ERAs currently would benefit broadly from the assay. This is not to say, however, that X. laevis might not be a valuable test species in other toxicological risk-assessment scenarios. For example, there has been a recent emphasis on the use of Xenopus as a model species for identifying chemicals with the potential to affect thyroid function in vertebrates ([96]; [97]). In addition, because X. laevis (and Xenopus tropicalis) can be cultured in the laboratory through a full life cycle, this species potentially could be a model for assessing the reproductive effects of contaminants in higher-tier risk assessments concerned specifically with amphibian reproduction [98].


  1. Top of page
  2. Abstract
  8. Acknowledgements

We thank J. Tietge and D. Mount for helpful discussions concerning various aspects of FETAX; S. Hunter and S. Degitz provided helpful comments on an earlier version of the manuscript. The information in this document has been funded in part by the U.S. Environmental Protection Agency. It has been subjected to review by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.


  1. Top of page
  2. Abstract
  8. Acknowledgements
  • 1
    Dumont J, Schultz TW, Buchanan M, Kao G. 1983. Frog embryo teratogenesis assay-Xenopus (FETAX)—a short-term assay application to complex environmental mixtures. In WatersMD, SandhuSS, LewtasJ, ClaxtonL, ChernoffN, NesnowS, eds, Short-Term Bioassays in the Analysis of Complex Environmental Mixtures, Vol 3. Plenum, New York, NY, USA, pp 393405.
  • 2
    Deuchar EM. 1972. Xenopus laevis and developmental biology. Biol Rev 47: 37112.
  • 3
    Deuchar EM. 1975. Xenopus: The South African Clawed Frog. John Wiley, New York, NY, USA.
  • 4
    American Society for Testing and Materials. 1991. Standard guide for conducting the frog embryo teratogenesis assay-Xenopus (FETAX). E1439–91. In Annual Book of ASTM Standards, Vol 11. Philadelphia, PA, pp 863873.
  • 5
    American Society for Testing and Materials. 1998. Standard guide for conducting the frog embryo teratogenesis assay-Xenopus (FETAX). E1439–98. In Annual Book of ASTM Standards, Vol 11. Philadelphia, PA, pp 116.
  • 6
    Bantle JA, Burton DT, Dawson DA, Dumont JN, Finch RA, Fort DJ, Linder G, Rayburn JR, Buchwalter D, Maurice MA, Turley SD. 1994. Initial interlaboratory validation study of FETAX: Phase I testing. J Appl Toxicol 14: 213223.
  • 7
    Bantle JA, Burton DT, Dawson DA, Dumont JN, Finch RA, Fort DJ, Linder G, Rayburn JR, Buchwalter D, Gaudet-Hull AM, Maurcie MA, Turley SD. 1994. FETAX interlaboratory validation study: Phase II testing. Environ Toxicol Chem 13: 16291637.
  • 8
    Bantle JA, Finch RA, Burton DT, Fort DJ, Dawson DA, Linder G, Rayburn JR, Hull M, Kumsher-King M, Gaudet-Hull AM, Turley SD. 1996. FETAX interlaboratory validation study: Phase III—part 1 testing. J Appl Toxicol 16: 517528.
  • 9
    Bantle JA, Finch RA, Fort DJ, Stover EL, Hull M, Kumsher-King M, Gaudet-Hull AM. 1999. Phase III interlaboratory study of FETAX, Part 3. FETAX validation using 12 compounds with and without an exogenous metabolic activation system. J Appl Toxicol 19: 447472.
  • 10
    Fort DJ, Stover EL, Bantle JA, Rayburn JR, Hull MA, Finch RA, Burton DT, Turley SD, Dawson DA, Linder G, Buchwalter D, Dumont JN, Kumsher-King M, Gaudet-Hull AM. 1998. Phase III interlaboratory study of FETAX, Part 2: Interlaboratory validation of an exogenous metabolic activation system for frog embryo teratogenesis assay Xenopus (FETAX). Drug Chem Toxicol 21: 114.
  • 11
    Bantle JA, Dawson DA. 1988. Uninduced rat liver microsomes as an in vivo metabolic activation system for the frog embryo teratogenesis assay-Xenopus (FETAX). In AdamsWJ, ChapmanGA, LandisWG, eds, Aquatic Toxicology and Hazard Assessment, Vol 19. STP 971. American Society for Testing and Materials, Philadelphia, PA, pp 316326.
  • 12
    Fort DJ, Dawson DA, Bantle JA. 1988. Development of a metabolic activation system for the Frog Embryo Teratogenesis Assay: Xenopus (FETAX). Teratog Carcinog Mutagen 8: 251263.
  • 13
    Dumont JN, Schultz TW. 1980. Effects of coal-gasification sour water on Xenopus laevis embryos. J Environ Sci Health A 15: 127138.
  • 14
    Dawson DA, McCormick CA, Bantle JA. 1985. Detection of teratogenic substances in acidic mine water samples using the frog embryo teratogenesis assay Xenopus (FETAX). J Appl Toxicol 5: 234244.
  • 15
    Dawson DA, Stebler E, Burks SA, Bantle JA. 1988. Evaluation of the developmental toxicity of metal-contaminated sediments using short-term fathead minnow and frog embryo-larval assays. Environ Toxicol Chem 7: 2734.
  • 16
    Bantle JA, Fort DJ, James BL. 1989. Identification of developmental toxicants using the frog embryo teratogenesis assay Xenopus (FETAX). Hydrobiologia 188/189: 577585.
  • 17
    Fort DJ, Stover EL, Norton D. 1995. Ecological hazard assessment of aqueous soil extracts using FETAX. J Appl Toxicol 15: 183191.
  • 18
    Fort DJ, Propst TL, Stover EL, Helgen JC, Levey RB, Gallagher K, Burkhart JG. 1999. Effects of pond water, sediment, and sediment extracts from Minnesota and Vermont, USA, on early development and metamorphosis of Xenopus. Environ Toxicol Chem 18: 23052315.
  • 19
    Burkhart JG, Helgen JC, Fort DJ, Gallagher K, Bowers D, Propst TL, Gernes M, Magner J, Shelby MD, Lucier G. 1998. Induction of mortality and malformation in Xenopus laevis embryos by water sources associated with field frog deformities. Environ Health Perspect 106: 841848.
  • 20
    Bantle JA. 1995. FETAX—a developmental toxicity assay using frog embryos. In RandGM, ed, Fundamentals in Aquatic Toxicology. Taylor & Francis, Washington, DC, USA, pp 207230.
  • 21
    Fort DJ, Stover EL. 1996. Assessing ecological hazard to amphibian populations. In Proceedings, 51st Purdue Industrial Waste Conference. Ann Arbor Press, Chelsea, MI, USA, pp 351358.
  • 22
    Blaustein AR, Wake DB. 1990. Declining amphibian populations: A global phenomenon? Trends Ecol Evol 5: 203204.
  • 23
    Wake DB. 1991. Declining amphibian populations. Science 253: 860.
  • 24
    Pechman JHK, Scott DE, Semlitsch RD, Caldwell JP, Vitt LJ, Gibbons JW. 1991. Declining amphibian populations: The problem of separating human impacts from natural fluctuations. Science 253: 892895.
  • 25
    Pechmann JHK, Wilbur HM. 1994. Putting declining amphibian populations in perspective: Natural fluctuations and human impacts. Herpetologica 50: 6584.
  • 26
    Corn PS. 1994. What we know and don't know about amphibian declines in the West. In Covington WW, Debano LF, Technical Coordinators, Sustainable Ecological Systems: Implementing an Ecological Approach to Land Management. General Technical Report RM-247. U.S. Department of Agriculture Forest Service, Fort Collins, CO, pp 5967.
  • 27
    Corn PS. 2000. Amphibian declines: Review of some current hypotheses. In SparlingDW, BishopCA, LinderG, eds, Ecotoxicology of Amphibians and Reptiles. SETAC, Pensacola, FL, USA, pp 663696.
  • 28
    McCoy ED. 1994. “Amphibian decline”: A scientific dilemma in more ways than one. Herpetologica 50: 98103.
  • 29
    Stebbins RC, Cohen NW. 1995. A Natural History of Amphibians. Princeton University, Princeton, NJ, USA.
  • 30
    GreenDM, ed. 1997. Herpetological Conservation, Vol 1. Society for the Study of Amphibians and Reptiles, Saint Louis, MO, USA.
  • 31
    Green DM. 1997. Perspectives on amphibian population declines: Defining the problem and searching for answers. In GreenDM, ed, Herpetological Conservation, Vol 1. Society for the Study of Amphibians and Reptiles, Saint Louis, MO, USA, pp 291308.
  • 32
    Ouellet M, Bonin J, Rodrique J, DesGranges J, Lair L. 1997. Hindlimb deformities (ectomelia, ectrodactyla) in free-living anurans from agriculture habitats. J Wildl Dis 33: 95104.
  • 33
    Gardiner DM, Hoppe DM. 1999. Environmentally induced limb malformation in mink frogs (Rana septentrionalis). J Exp Zool 284: 207216.
  • 34
    Degitz SJ, Kosian PA, Makynen EA, Jensen KM, Ankley GT. 2000. Stage- and species-specific developmental toxicity of alltrans retinoic acid in four native North American ranids and Xenopus laevis. Toxicol Sci 57: 264274.
  • 35
    Degitz SJ, Durhan EJ, Tietge JE, Kosian PA, Holcombe GW, Ankley GT. 2003. Developmental toxicity of methoprene and several degradation products to Xenopus laevis. Aquat Toxicol 64: 97105.
  • 36
    Boyer R, Grue CE. 1995. The need for water quality criteria for frogs. Environ Health Perspect 103: 352357.
  • 37
    Carey C, Bryant CJ. 1995. Possible interrelations among environmental toxicants, amphibian development, and decline of amphibian populations. Environ Health Perspect 103: 1317.
  • 38
    McKim JM. 1977. Evaluation of tests with early life stages of fish for predicting long-term toxicity. J Fish Res Board Can 34: 11481154.
  • 39
    U.S. Environmental Protection Agency. 1994. Short-term methods for estimating the chronic toxicity of effluents and receiving waters to freshwater organisms, 3rd ed. EPA 600/4–91–002. Duluth, MN.
  • 40
    U.S. Environmental Protection Agency. 1994. Short-term methods for estimating the chronic toxicity of effluents and receiving waters to marine and estuarine organisms. EPA 600/4–91–003. Cincinnati, OH.
  • 41
    Organization for Economic Cooperation and Development. 1992. Guidelines for testing of chemicals. Section 2, Guideline 204. Fish, prolonged toxicity test: 14-day study. Paris, France.
  • 42
    Organization for Economic Cooperation and Development. 1992. Guidelines for testing of chemicals. Section 2, Guideline 210. Fish early life-stage toxicity test. Paris, France.
  • 43
    U.S. Environmental Protection Agency. 2000. Methods for measuring the toxicity and bioaccumulation of sediment-associated contaminants with freshwater invertebrates, 2nd ed. EPA 600/R99/064. Duluth, MN.
  • 44
    Brown LE, Rosati RR. 1997. Effects of three different diets on survival and growth of larvae of the African clawed frog Xenopus laevis. Progressive Fish Culturist 59: 5458.
  • 45
    Green SL. 2002. Factors affecting oogenesis in the South African clawed frog (Xenopus laevis). Comparative Medicine 52: 307312.
  • 46
    Woods PE, Paulauskis JD, Weigt LA, Romano MA, Guttman SI. 1989. Genetic variation of laboratory and field populations of the midge, Chironomus tentans Fab.: Implications for toxicology. Environ Toxicol Chem 8: 10671074.
  • 47
    Diamond SA, Newman MC, Mulvey M, Guttman SI. 1991. Allozyme genotype and time-to-death to mosquitofish, Gambusia holbrooki, during acute inorganic mercury exposure: A comparison of populations. Aquat Toxicol 21: 119134.
  • 48
    Guttman SI. 1994. Population genetic structure and ecotoxicology. Environ Health Perspect 102: 97100.
  • 49
    Duan Y, Guttman SI, Oris JT. 1997. Genetic differentiation among laboratory populations of Hyalella azteca: Implications for toxicology. Environ Toxicol Chem 16: 691695.
  • 50
    Hazel JR, Prosser CL. 1974. Molecular mechanisms of temperature compensation in poikilotherms. Physiol Rev 54: 620677.
  • 51
    Hochachka PW, Somero GN. 1973. Strategies of Biochemical Adaptation. W.B. Saunders, Philadelphia, PA, USA.
  • 52
    Nieuwkoop PD, Faber J. 1994. Normal Tables of Xenopus laevis (Daudin), A Systematical and Chronological Survey of the Development from the Fertilized Egg till the End of Metamorphosis, 3rd ed. Garland, New York, NY, USA.
  • 53
    American Society for Testing and Materials. 1991. Standard guide for conducting acute toxicity tests with fishes, macroinvertebrates, and amphibians. E729–88a. In Annual Book of ASTM Standards, Vol 11. Philadelphia, PA, USA, pp 249268.
  • 54
    Schuytema GS, Nebeker AV, Griffis WL. 1994. Toxicity of Guthion and Guthion 2S to Xenopus laevis embryos. Arch Environ Contam Toxicol 27: 250255.
  • 55
    Fort DJ, Stover EL. 1997. Significance of experimental design in evaluating ecological hazards of sediments/soils to amphibian species. In DwyerFT, DoansTR, HinmanML, eds, Environmental Toxicology and Risk Assessment: Modeling and Risk Assessment, Vol 6. STP 1317. American Society for Testing and Materials, Philadelphia, PA, pp 427442.
  • 56
    Dawson DA, Bantle JA. 1987. Development of a reconstituted water medium and preliminary validation of the frog embryo teratogenesis assay Xenopus (FETAX). J Appl Toxicol 7: 237244.
  • 57
    Tietge JE, Ankley GT, DeFoe DL, Holcombe GW, Jensen KM. 2000. Effects of water quality on development of Xenopus laevis: A FETAX assessment of surface water associated with malformations in native anurans. Environ Toxicol Chem 19: 21142121.
  • 58
    Schuytema GS, Nebeker AV. 1999. Comparative effects of ammonium and nitrate compounds on Pacific treefrog and African clawed frog embryos. Arch Environ Contam Toxicol 36: 200206.
  • 59
    U.S. Environmental Protection Agency. 1985. Ambient water quality criteria for ammonia. EPA 440/05–85/001. Washington, DC.
  • 60
    Garber EA. 2002. Mineral deficiency and the use of the FETAX bioassay to study environmental teratogens. J Appl Toxicol 22: 237240.
  • 61
    Luo SQ, Plowman MC, Hopfer SM, Sunderman FW Jr. 1993. Mg+2-deprivation enhances and Mg+2-supplementation diminishes the embryotoxic and teratogenic effects of Ni+2, Co+2, Zn+2, Cd+2 for frog embryos in the FETAX assay. Ann Clin Lab Sci 23: 121129.
  • 62
    U.S. Environmental Protection Agency. 1985. Guidelines for deriving numerical national water quality criteria for the protection of aquatic organisms and their uses. EPA 822/R-85–100. Washington, DC.
  • 63
    U.S. Environmental Protection Agency. 1989. Pesticide assessment guidelines, subdivision E, hazard evaluation: Wildlife and aquatic organisms. EPA 540/09–82–024. Washington, DC.
  • 64
    U.S. Environmental Protection Agency. 1996. OPPTS test guidelines. Series 850: Ecological effects, Vol 1, Public Draft. Washington, DC.
  • 65
    Fort DJ, James BL, Bantle JA. 1989. Evaluation of the developmental toxicity of five compounds using the frog embryo teratogenesis assay: Xenopus and a metabolic activation system. J Appl Toxicol 9: 377388.
  • 66
    Guengerich FP. 1979. Isolation and purification of cytochrome P-450 and the existence of multiple forms. Pharmacol Ther 6: 99121.
  • 67
    Bantle JA, Dumont JN, Finch RA, Linder G. 1991. Atlas of Abnormalities: A Guide for the Performance of FETAX. Oklahoma State University, Stillwater, OK, USA.
  • 68
    Gotleb AC, Appelman J, Bronkhorst MC, van den Berg JHJ, Spenkelink A, Brower A, Murk AJ. 1999. Delayed effects of preand early lifetime exposure to polychlorinated biphenyls on tadpoles of two amphibian species (Xenopus laevis and Rana temporaria). Environ Toxicol Pharmacol 8: 114.
  • 69
    Harris ML, Bishop CA, McDaniel TV. 2001. Assessment of rates of deformity in wild frog populations using in situ cages: A case study of leopard frogs (Rana pipiens) in Ontario, Canada. Biomarkers 6: 5263.
  • 70
    Schoff PK, Johnston CM, Schotthoefer AM, Murphy JE, Lieske C, Cole RA, Johnson LB, Beasley VR. 2002. Estimations of the prevalence of malformed frogs. Abstracts, Symposium on Multiple Stressor Effects in Relation to Declining Amphibian Populations, April 16–17, American Society for Testing and Materials International, Philadelphia, PA, p 19.
  • 71
    Morgan MK, Scheuerman PR, Bishop CS, Pyles RA. 1996. Teratogenic potential of atrazine and 2,4-D using FETAX. J Toxicol Environ Health 48: 151168.
  • 72
    Osano O, Oladimeji AA, Kraak MHS, Admiraal W. 2002. Teratogenic effects of amitraz, 2,4-dimethylaniline, and paraquat on developing frog (Xenopus) embryos. Arch Environ Contam Toxicol 43: 4249.
  • 73
    Osano O, Admiraal W, Otien D. 2002. Developmental disorders in embryos of the frog Xenopus laevis induced by chloroacetanilide herbicides and their degradation products. Environ Toxicol Chem 21: 375379.
  • 74
    Schuytema GS, Nebeker AV. 1996. Amphibian toxicity data for water quality criteria chemicals. EPA/600/R-96/124. U.S. Environmental Protection Agency, Corvallis, OR.
  • 75
    Birge WJ, Black JA, Westerman AG. 1979. Evaluation of aquatic pollutants using fish and amphibian eggs as bioassay organisms. In NielsenSW, MigakiG, ScarpelliDG, eds, Animals as Monitors of Environmental Pollutants. National Academy of Sciences, Washington, DC, USA.
  • 76
    Linder G, Barbitta J, Kwaiser T. 1990. Short-term amphibian toxicity tests and paraquat toxicity assessment. In Landis WG, Van der Schalie WH, eds, Aquatic Toxicology and Risk Assessment, 13th Symposium. STP 1096. American Society for Testing and Materials, Philadelphia, PA, pp 189198.
  • 77
    Birge WJ, Westerman AG, Spromsberg JA. 2000. Comparative toxicology and risk assessment of amphibians. In SparlingDW, LinderG, BishopCA, eds, Ecotoxicology of Amphibians and Reptiles. SETAC, Pensacola, FL, USA, pp 727791.
  • 78
    U.S. Environmental Protection Agency. 1980. Ambient water quality criteria for aldrin/dieldrin. EPA 440/5–80–019. Office of Water, Washington, DC.
  • 79
    U.S. Environmental Protection Agency. 1985. Ambient water quality criteria for cadmium. EPA 440/5–84–032. Office of Water, Washington, DC.
  • 80
    U.S. Environmental Protection Agency. 1985. Ambient water quality criteria for copper. EPA 440/5–84–031. Office of Water, Washington, DC.
  • 81
    U.S. Environmental Protection Agency. 1985. Ambient water quality criteria for mercury. EPA 440/5–84–026. Office of Water, Washington, DC.
  • 82
    U.S. Environmental Protection Agency. 1985. Ambient water quality criteria for chromium. EPA 440/5–84–029. Office of Water, Washington, DC.
  • 83
    U.S. Environmental Protection Agency. 1985. Ambient water quality criteria for nickel. EPA 440/05–85–060. Office of Water, Washington, DC.
  • 84
    U.S. Environmental Protection Agency. 1986. Ambient water quality criteria for pentachlorophenol. EPA 440/5–86–009. Office of Water, Washington, DC.
  • 85
    U.S. Environmental Protection Agency. 1986. Ambient water quality criteria for parathion. EPA 440/5–86–007. Office of Water, Washington, DC.
  • 86
    U.S. Environmental Protection Agency. 1987. Ambient water quality criteria for selenium. EPA 440/5–87–006. Office of Water, Washington, DC.
  • 87
    U.S. Environmental Protection Agency. 1987. Ambient water quality criteria for zinc. EPA 440/05–80–079. Office of Water, Washington, DC.
  • 88
    U.S. Environmental Protection Agency. 1988. Ambient water quality criteria for aluminum. EPA 440/5–86–008. Office of Water, Washington, DC.
  • 89
    U.S. Environmental Protection Agency. 1993. Ambient aquatic life water quality criteria for aniline. Draft 9/22/93, Office of Water, Washington, DC.
  • 90
    Birge WJ, Black JA, Westerman AG. 1985. Short-term fish and amphibian embryo-larval tests for determining the effects of toxicant stress on early life stages and estimating chronic values for single compounds and complex effluents. Environ Toxicol Chem 4: 807821.
  • 91
    Fort DJ, Stover EL, Bantle JA. 1996. Integrated ecological hazard assessment of waste site soil extracts using FETAX and short-term fathead minnow teratogenesis assay. In LaPointTW, PriceFT, LittleEE, eds, Environmental Toxicology and Risk Assessment, Vol 4. STP 1262. American Society for Testing and Materials, Philadelphia, PA, pp 93109.
  • 92
    Schuytema GS, Nebeker AV, Griffis WL, Wilson KN. 1991. Teratogenesis, toxicity, and bioconcentration in frogs exposed to dieldrin. Arch Environ Contam Toxicol 21: 332350.
  • 93
    Jofre MB, Karasov WH. 1999. Direct effect of ammonia on three species of North American anuran amphibians. Environ Toxicol Chem 18: 18061812.
  • 94
    Bridges CM, Dwyer FJ, Hardesty DK, Whites DW. 2002. Comparative contaminant toxicity: Are amphibian larvae more sensitive than fish? Bull Environ Toxicol Chem 69: 562569.
  • 95
    Hall RJ, Henry PFP. 1992. Assessing effects of pesticides on amphibians and reptiles: Status and needs. Herpetological Journal 2: 6571.
  • 96
    U.S. Environmental Protection Agency. 1998. Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC) Report. Office of Pollution Prevention and Toxic Substances, Washington, DC.
  • 97
    Opitz R, Braunbeck T, Bogi C, Pickford DB, Nentwig G, Oehlmann J, Tooi O, Lutz I, Kloas W. 2005. Description and initial evaluation of a Xenopus metamorphosis assay for detection of thyroid system-disrupting activities of environmental chemicals. Environ Toxicol Chem 24: 663664.
  • 98
    Ankley G, Mihaich E, Stahl R, Tillitt D, Colborn T, McMaster S, Miller R, Bantle J, Campbell P, Denslow N, Dickerson R, Folmar L, Fry M, Giesy J, Gray LE, Guiney P, Hutchinson T, Kennedy S, Kramer V, LeBlanc G, Mayes M, Nimod A, Patino R, Peterson R, Purdy R, Ringer R, Thomas P, Touart L, Van Der Kraak G, Zacharewski T. 1998. Overview of a workshop on screening methods for detecting potential (anti-) estrogenic/androgenic chemicals in wildlife. Environ Toxicol Chem 17: 6887.
  • 99
    Fort DJ, Stover EL. 1996. Effect of low-level copper and pentachlorophenol exposure on various early life stages of Xenopus laevis. In BengstonDA, HenshelDS, eds, Environmental Toxicology and Risk Assessment: Biomarkers and Risk Assessment, Vol 5. STP 1306. American Society for Testing and Materials, Philadelphia, PA, pp 188203.