Ecological Risk Assessment for Reptiles and Amphibians
Comparative assessment of methodologies and available data
Mark S. Johnson (WORKING DRAFT)
Abstract
Historically, ecological risk assessments have rarely included amphibian and reptile species and have focused on other aquatic (fish, copepods, algae) and terrestrial wildlife (birds and mammal) species. Often this lack of consideration is due to a paucity of toxicity data, significant variation in study design, or a combination of both. Guidance is needed in the assessment of appropriate methods, development of useful laboratory animal models, and an examination of biological patterns in response to develop guidance to further the goals in protecting and more accurately describing risk from exposures to of xenobiotics to reptiles and amphibians. Patterns in toxicity may help to ascertain whether results from some classes of vertebrates could be representative of others (e.g. whether fish are appropriate surrogates for aquatic amphibian life stages) in hopes of refining testing regimes. This scope is intended to be a guide in the development of methods that would yield data appropriate for ecological risk decisions applicable to amphibians and reptiles.
Introduction: Ecological risk assessment can be categorized into two general categories, a priori or post hoc. Examples of a priori assessment include pesticide registration, incinerator or waste water permits, or other regulatory requirements that include the prediction of exposure criteria that are then compared to threshold toxicity values (e.g. Toxicity Reference Values or risk-based media concentrations) to make risk assessment decisions. Since environmental contamination has not yet occurred, various assumptions are made to infer on the magnitude of exposure.
Post hoc risk assessments provide information that allow for decisions to be made at contaminated sites. Here, exposure can be measured, through analytical chemistry of contaminated media and/or through measures of species-specific body burdens and assessment of biomarkers. Measures of effect may also be evaluated.
Risk assessment data requirements for each of the two risk assessment forms can be quite different, where a priori assessments will strive to test toxicity using pure parent compound in a controlled laboratory environment using a species that has the potential for a low probability of false negatives (low potential for Type I error); post hoc risk assessments often begin with the use of controlled laboratory data, but then have the additional option of collecting field data to refine screening- level risk assessment predictions. Regardless, there is value in reviewing and defining methods used and identification of appropriate laboratory animal models for reptile and amphibian species for both forms of risk assessment applications.
Laboratory animal models for amphibians and reptiles are much more limited than those for other classes of vertebrates. Differences in life history, physiology, and relative influence of environmental variables suggest a focused approach in the valuation of specific models in controlled testing regimes is needed. As biologists search for patterns in toxicity that may be associated with physiological, environmental, or behavioral attributes, a refined review of the literature in regards to risk assessment has not been done. This effort attempts to help provide guidance and background information to support this goal.
Consistent with these goals are several questions for the two types of risk assessment categories described earlier.
A priori and Post hoc – Laboratory model development
Amphibians – a) Are fish appropriate surrogates (low Type I error)?
b) Which species make appropriate laboratory models?
c) What established methods exist for toxicity testing?
d) What in situ methods exist for amphibians (for post hocexposures)?
e) How are exposure routes integrated for amphibians and which are predominant (e.g. oral vs. dermal)?
f) Which species/genera show increased sensitivity to which class of compounds? Are there patterns in sensitivity related to metamorphic transitions that most pronounced in some species? What substances are most likely to affect metamorphosis?
Reptiles - a) Are data from avian models representative for reptile toxicity ?
b) Which species are appropriate laboratory models?
c) What established methods exist for toxicity testing?
Each of these questions is explored as follows.
Are fish appropriate surrogates?
Several studies have comparatively reviewed the relative toxicity of compounds in a water matrix to exposures in fish and amphibian species with mixed results. Birge et al. (2000) provided a review of results from acute early-life-stage tests to 25 amphibian species compared with data for rainbow trout (Oncorhynchusmkiss). This comparison included test results from 34 inorganic, 27 organic compounds and based on 447 embryo-larval tests with amphibian and fish species. LC50 and LC10 values were calculated for combined test responses by log probit analysis. A chemical hazard index (CHI) value was calculated using the LC50 amphibian value/rainbow trout LC50 where a CHI <1 indicated increased tolerance compared to trout and a CHI >1 increased sensitivity. Frogs for the genus Hyla and Rana(Lithobates) trended to be most sensitive of the compounds tested; amphibians were more sensitive than were rainbow trout in 35% of comparisons with organic compounds. Toads of the genus Bufo trended to be more tolerant than other amphibian species tested (Birge et al. 2000). Overall, amphibian species had lower mean CHI values than rainbow trout in 52% of the comparisons with metals and 49% for all chemicals. Additionally, comparisons were made between species for specific chemicals.
Weltje et al. (2013) evaluated data from the USEPA ECTOX database and other sources and evaluated the acute and chronic data for amphibians and fish and found similar patterns in relative toxicity. Acute data used LC50 values and chronic data included No Observed Effect Concentrations (NOECs) that followed specific acceptability criteria (e.g. were bounded by LOECs, not limit values, had treatments that were ≤ 100 between concentrations). Ninety –six hour LC50 comparisons of amphibian and fish for eight inorganic and 47 organic chemicals were made. Here, the median sensitivity ratio was 0.52 implying that overall fish were slightly more sensitive than amphibians; however, amphibians were more sensitive in 16 or 55 cases. Amphibians were between 10 and 100-fold more sensitive for four substances (2,4-dichlorophenoxyacetic acid, aluminum chloride, malathion, and pentachlorophenol); two amphibians were more than 100-fold more sensitive (dimethoate and p-nonylphenol).
Chronic comparisons included 52 chemicals including 10 inorganic and 42 organic that had sufficient fish and amphibian data. Here, the median sensitivity ratio was 0.21 implying that overall the fish were more sensitive than amphibians. Amphibians were more sensitive than fish in 11 of 52 cases; amphibians were between 10 and 100-fold more sensitive in five cases and more than 100-fold more sensitive in one case (sodium perchlorate, a thyroid/iodide inhibitor). A complicating factor in this analysis involves the use of NOECs and LOECs in these comparisons (Weltje et al. 2013). The authors recognized the increased sensitive to amphibians for phenolic compounds and those that may affect metamorphosis (e.g. thyroid inhibitors; those that act through the hypothalamic-pituitary-thyroid axis). Weltje et al. (2013) suggest that amphibians are not particularly more sensitive than fish in many cases and were within two orders of magnitude – the level of protection applied to fish data for European regulatory purposes. Other authors have found similar trends (Kerby et al. 2010, Bridges et al. 2002, De Young et al. 1996) though their database of substances was limited and sparse in any cases. Detailed examinations using dose-response functions/relationship (rather than point estimates) are needed for a more accurate comparison regarding sensitivities between fish and amphibian species. Further, few data are available for terrestrial amphibian species and a comparison of sediment/soil tests have not yet been evaluated.
(MORE NEEDED)
Which species make appropriate laboratory models?
Successful laboratory models for testing contaminated water and sediment included many larval Ranid species (e.g. Ranapipiens, R. catesbienna) with the most common species used testing compounds in water was Xenopuslaevis which was standardized as a developmental toxicity screen in the FETAX assay (Bantle et al. 1994). Most models use amphibian larvae in water exposures. Many species can be successfully reared from fertilized egg through metamorphosis, though the proper husbandry of adult frogs can be problematic in provisioning of carnivorous adult stage anurans. Salamander species that can be used successfully as laboratory models as adults include Plethodoncinereus, Ambystomamaculatum, Ambystomatigrium, and Ambystommexicanum, the latter used successfully for years from an in-bred population used in medical research (Axolotyl). Plethodon species can be successfully maintained in laboratory environment using Drosophila as food (Jeager1992, Johnson et al. 2004);Ambystomaspecies readily accept Lumbriculus earthworm species (Johnson et al. 2000). Here, oral estimates of dose can be quantified; however, toxicity of the compound to live food items is an important factor in study design.This design works best for compounds that have a high potential for biomagnification, yet relatively low toxicity to invertebrates.
(MORE NEEDED)
What established methods exist for toxicity testing?
Many established methods exist in OECD, USEPA, and ASTM formats. Other methods have been described in the literature.
USEPA 1996.Tadpole/Sediment Subchronic Toxicity Test. OPPTS 850.1800, EPA 712-C-96-132.Office of Prevention, Pesticides and Toxic Substances, Washington D.C.
OECD 231: Amphibian Metamorphosis Assay (2009) ISBN: 9789264076242.
ASTM E-2591 – Standard Guide for Conducting Whole Sediment Toxicity Tests with Amphibians.
ASTM E-729 – Standard Guide for Conducting Acute Toxicity Tests on Test Materials with Fishes, Macroinvertebrates and Amphibians.
ASTM E-1192 – Standard Guide for Conducting Acute Toxicity Tests on Aqueous Ambient Samples and Effluent with Fishes, Macrointervebrates, and Amphibians.
(MORE NEEDED; a table with citations would be useful)
What in situ methods exist for amphibians (for post hoc exposures)?
Methods to test contaminated media can include many adapted from methods described previously. Others include:
ASTM E-2591 – Standard Guide for Conducting Whole Sediment Toxicity Tests with Amphibians.
USEPA 1996.Tadpole/Sediment Subchronic Toxicity Test. OPPTS 850.1800, EPA 712-C-96-132.Office of Prevention, Pesticides and Toxic Substances, Washington D.C.
(MORE NEEDED)
How are exposure routes integrated for amphibians and which are predominant (e.g. oral vs. dermal)?
Many holistic exposure methods include contaminated media, but not contaminated food. Some studies have shown dermal exposures can contribute significantly to total exposure (Johnson et al. 1999). Testing guidelines are needed for compounds that have a potential to be released to soil, using dermal flux parameters to consider whether in situ soil exposure regimes should be included in data collection. Further, the relative importance of the oral route should also be assessed to determine whether test design should include oral exposures.
(MORE NEEDED)
Which species/genera show increased sensitivity to which class of compounds?
Are there patterns in sensitivity related to metamorphic transitions that most pronounced in some species? What substances are most likely to affect metamorphosis?Hypothalamus-pituitary-thyroid axis that can influence metamorphosis? Phenolics? Soil (dermal) exposures? Are some species more sensitive than others (see Birge et al. 2000; Hylids?)?
Which endpoints are inadequately evaluated in current testing regimes?
Adults allow for a more rigorous evaluation of biological processes. Some have begun to characterize immune functions (e.g. phagocytosis and ROI production of splenocytes), blood parameters (PCV, RBC/WBC counts, five-part differentials), and histopathology (see Johnson et al. 2000). There is a need to expand beyond the typical gross endpoints of mortality, growth, reproduction, and timing to development if we are to better understand toxicant/class interactions.
Are data from avian models representative for reptile toxicity ?
Briefly, few comparisons/reviews have been conducted evaluating whether data from birds could be extrapolated to protect reptilian species. Differences in physiology and cellular processes in species within each vertebrate class suggest relationships are likely extraneous at best. Fryday and Thompson (2009) conducted a review of the literature and found few useful data to make comparisons. Available data suggest that birds are typically sensitive to neurological toxicants, though responses and methods for useful comparisons are variable. Differences between species in metabolic requirements (e.g. food consumption which affects exposure) may be a consequence of internal body temperature regulation. Talent (2005) found that dermal exposure to pyrethrins was more toxic when administered to Anoliscarolinensis at 15-20⁰C than when maintained between 35-38⁰C. Others have found more complex interactions (McFarland et al. 2012). Regardless, a comparative review as are standardized methods are needed as more data become available.
(MORE NEEDED)
Which reptile species are appropriate laboratory models?
Several reptile species have been used in the laboratory with successful results. Western fence lizards (Sceloporusoccidentalis) have been captive bred for laboratory toxicity investigations to characterize acute, subacute and subchronic oral exposures (Talent et al. 2002, Suski et al. McFarland et al. 2009). Other investigations have included Anoline lizards (Talent 2005) and snapping turtles….
(MORE NEEDED)
What established methods for reptiles exist for toxicity testing?
No established/harmonized methods exist for reptiles, though various methods exist in the scientific literature, making toxicological comparisons of the data difficult (Fryday and Thompson 2009). Clearly, standardized methods and models are needed to be able to provide data for useful comparisons between species and vertebrate classes.
(MORE NEEDED)
What endpoints should be investigated as part of a standardized testing regime for reptiles?
(Suggest stagewiseprobit for acute testing to develop slope for comparison purposes, 14-d oral subacute, >6 treatments, then 60-d oral subchronic investigating suspected targets (e.g. histopathology or main blood conditioning organs, PCV, RBC/WBC counts, WBC 5-part diff, t-prot, glu, other clinical chemistries as indicated). Use benchmark dose tocharacterize dose response relationships (not EC-values or NOAELS/LOAEL relationships).
CONCLUSIONS
REFERENCES
Bantle JA, Rayburn JR, Burton DT, Turley SD, Dawson DA, Dumont JN, Finch RTA, Maurice MA, Fort DJ, Linder G, Buchwalter D. 1994. Initial interlaboratory validation study of FETAX: Phase I testing. J. Appl. Toxicol. 14:213-223.
Birge WJ, Westerman AG, Spromberg JA. 2000. Comparative toxicology and risk assessment of amphibians. Pp. 727-791 in Ecotoxicology of Amphibians and Reptiles (Sparling DW, Linder G, and Bishop CA, eds). SETAC Press, Pensacola, FL, USA.
Bridges CM, Dwyer FJ, Hardesty DK, Whites DW. 2002. Comparative contamninant toxicity: are amphibian larvae more sensitive than fish? Bull. Environ. Contam.Toxicol. 69:562-569.
DeYoung DJ, Bantle JA, Hull MA, Burks SL. 1996. Differences in sensitivity to developmental toxicants as seen in Xenopus and Pimphalesembyros. Bull. Environ. Contam.Toxicol. 56:143-150.
Fryday S, Thompson H. 2009. Comparative toxicity of chemicals to reptiles and other vertebrates. A Report to EFSA CFT/EFSA/PPR/2008/01 Lot 2, Environmental Risk Assessment Team, Environmental Risk Programme, The Food and Environmental Research Agency, York YO41 1LZ, UK.
Jaeger RG. 1992. Housing, handling, and nutrition of salamanders. On: Schaeffer DO, Kleinow KM, Krulisch L, eds) The Care and Use of Amphibians, Reptiles, and Fish in Research. Scientists Center for Animal Welfare, Bethesda, MD.
Johnson MS,Holladay SD, Lippenholz KS, Jenkins JL, McCain WC. 2000. Effects of 2,4,6-trinitrotoluene in a holistic environmental exposure regime to a terrestrial salamander: Ambystomatigrinum. Toxicol.Pathol. 28:334-341.
Johnson MS, Paulus HI, Salice CJ, Checkai RT, Simini M. 2004.Toxicologic and histopathologic response of the terrestrial salamander, Plethodoncinereus to soil exposures of 1,3,5-trinitrohexahydro-1,3,5-triazine. Arch. Environ. Contam.Toxicol. 47:496-501.
Kerby JL, Richards-Hrdlicka KL, Storfer A, Skelly DK. 2010. An examination of amphibian sensitivity to environmental contaminants: are ampohibians poor canaries? Ecol. Lett. 12:1-8.
McFarland CA, Talent LG, Quinn MJ Jr, Bazar MA, Wilbanks MS, Nisanian M, Gogal RM Jr, Johnson MS, Perkins EJ, Gust KA. 2012. Multiple environmental stressors elicit complex interactive effects in the western fence lizard (Sceloporusoccidentalis). Ecotoxicology 21:2372-2390.
McFarland CA, Quinn MJ Jr, Bazar MA, Talent LG, Johnson MS. 2009. Toxic effects of oral hexahydro-1,3,5-trinitro-1,3,5-triazine in the western fence lizard (Sceloporusoccidentalis). Environ. Toxicol.Chem. 28:1043-1050.
Suski JG, Salice C, Houpt JT, Bazar MA, Talent LG. 2008. Dose-related effects following oral exposure of 2,4-dinitrotoluene on the Western fence lizard, Sceloporusoccidentalis. Environ. Toxicol. Chem. 27:352-359.
Talent LG, Dumont JN, Bantle JA, Janz DM, Talent SG. 2002. Evaluation of Western fence lizards (Sceloporusoccidentalis) and Eastern fence lizards (Sceloporus undulates) as laboratory reptile models for toxicological investigations. Environ. Toxicol. Chem. 21:899-905.
Talent LG. 2005.Effect of temperature on toxicity of a natural pyrethrin pesticide to green anole lizards (Anoliscarolinensis). Environ. Toxicol. Chem. 24:3113-3116.
Weltje L, Simpson P, Gross M, Crane M, Wheeler JR. 2013. Comparative acute and chronic sensitivity of fish and amphibians: a critical review of data. Environ. Toxicool. Chem. 32:984-994.
1