SEARCH

SEARCH BY CITATION

A major uncertainty in many aquatic risk assessments for toxic chemicals is the aggregate effect of the physicochemical characteristics of exposure media on toxicity and how this affects extrapolation of laboratory test results to natural systems. A notable example of this is how metal toxicity in freshwater varies because of factors such as water hardness, alkalinity, pH, dissolved organic carbon, and suspended solids. This has been the subject of hundreds of studies over the last 50 yr and of various papers in Environmental Toxicology and Chemistry since its inception, including 4 of the “Top 100” cited papers [1-4]. One study found median lethal concentrations (LC50s) for acute copper toxicity to fathead minnows to vary by more than 100-fold across various exposure water compositions well within the range observed in natural systems [1]. Approaches for modeling and predicting such variation have also been the subject of these efforts. An important example of this is the biotic ligand model (BLM), an approach first described and implemented in Di Toro et al. [2] and Santore et al. [3], and a focus of considerable research activity and regulatory development over the last 15 years.

GENESIS OF THE BLM APPROACH

  1. Top of page
  2. GENESIS OF THE BLM APPROACH
  3. CHALLENGES FOR THE BLM APPROACH
  4. SUPPLEMENTAL DATA
  5. Acknowledgment
  6. REFERENCES
  7. Supporting Information

The BLM represents an integration of decades of work regarding metal speciation, accumulation, toxicity, and physiology [5] that can be only briefly summarized here. Studies in the 1960s and 1970s demonstrated that the toxicity of several metals to freshwater organisms varied with hardness, alkalinity, pH, and organic ligands, for which 2 major mechanisms were inferred [6-8]. First, the toxicity of a metal varies with its chemical speciation and often is closely correlated with the “free” (uncomplexed) metal ion, although in some situations other metal species contribute to accumulation or toxicity. Second, cations such as calcium and hydrogen ions can have effects on toxicity independent of toxic metal speciation; these cations were postulated to ameliorate toxicity by competing with toxic metal ions for binding sites on the gill [9]. These mechanisms pertain to bioavailability—how exposure conditions affect the amount of metal accumulation relative to the total concentration in the exposure medium. However, calcium was also recognized to have effects on gill permeability and thus could affect toxicity other than as a competing cation [6, 8].

Some of these early toxicological findings were incorporated into water quality regulations, such as the hardness dependence of the US Environmental Protection Agency's (USEPA) aquatic life criteria for metals [10]. However, such simple correlation models incompletely describe the effects of exposure chemistry. A broader, mechanistic perspective of these effects was articulated by Morel [11], and by Pagenkopf [12] in his gill surface interaction model (GSIM). In the GSIM, bioavailability is modeled based on the binding of toxic metal species (free metal or other species) to sites on the gill surface, this binding being affected by 1) metal speciation that determines the activity of the toxic metal species, and 2) other cations (e.g., calcium, hydrogen ion) that compete with the toxic metal for the binding sites. Toxicity is modeled as a simple, empirical correlate of this accumulation and, to the extent that toxicity involves internalization of the metal and transport to other sites, this is assumed to be proportional to accumulation at the gill surface. Playle and associates [13, 14] supplied key early support for this model by demonstrating toxic metal accumulation on fish gills to be reduced both by metal-complexing ligands and by other cations, and also by measuring key model parameters.

Regulatory application of such a model was hampered by the limited availability of suitable effects data for a variety of species and the lack of appropriate meta-analysis of available data. Also problematic was measuring or predicting metal speciation for the complex mixtures of organic matter in natural waters, although significant advances were also being made regarding this [5, 15]. The different elements of this metal toxicity issue came together at a SETAC Pellston workshop [16], where it was evident that existing approaches might address the various elements of a more comprehensive model suitable for regulatory application, although various uncertainties and challenges were also recognized [15].

With industry and regulatory agency support, efforts resulted in what was christened the BLM, which combined a GSIM-type effects model with a metal speciation model [2, 3]. The term biotic ligand was adopted to expand this modeling concept beyond gills, and also to emphasize how toxicologically relevant biochemical sites on and in an organism are ligands in competition for the toxic metal with ligands in the exposure media. In its strictest sense, the biotic ligand must be surficial to interact with the external media, but can be more broadly viewed as including internal sites whose accumulation is proportional to the surface accumulation.

A flurry of research activity in the late 1990s and early 2000s further developed the BLM approach and expanded its scope beyond the initial emphasis on acute toxicity to fish to other endpoints, taxa, and routes of exposure [5, 17, 18]. This had regulatory consequences, including incorporation of the BLM approach into the USEPA aquatic life criterion for copper [19] and other actions in various jurisdictions [20]. Significant BLM research and development continues.

A major strength of the biotic ligand concept is that it provides a focus for organizing information on how exposure water chemistry affects bioavailability (as represented by metal accumulation at toxicologically relevant sites) and on how toxicity results from this accumulation. This provides the possibility of parameterizing BLMs independently of any water-based effects concentrations, based on measured values for binding constants of the toxic metal and competing cations to the biotic ligand, the metal-binding capacity of the biotic ligand, and the accumulation on the biotic ligand associated with a particular toxic effect (e.g., an LA50 [the accumulation associated with 50% mortality]). Such parameterization allows independent predictions of water-based effects concentrations and evaluation of the mechanistic validity of the model, but adequate data in this regard are rarely available.

In practice, 1 or more model parameters are usually estimated by calibration to water-based effects concentrations. Although this provides weaker validation for mechanisms than predictions using accumulation-based parameterization, it still provides mechanistically based equations for the relationship of toxicity to exposure conditions. For example, Meyer [21] described BLM-based equations for the type of hardness relationships used in USEPA water quality criteria and how these model equations relate to the simple correlations actually in the criteria. Partially parameterizing the BLM with water-based effects concentrations actually has the benefit of reducing the number of parameters, because the toxic metal binding constant, the binding capacity, and the toxic accumulation all combine into a single term equal to a water-based effect concentration in the absence of complexing ligands and competing cations [18, 19]. This simplifies the BLM to a product of this limiting concentration, a factor related to toxic metal speciation, and a factor related to cation competition [18, 19].

CHALLENGES FOR THE BLM APPROACH

  1. Top of page
  2. GENESIS OF THE BLM APPROACH
  3. CHALLENGES FOR THE BLM APPROACH
  4. SUPPLEMENTAL DATA
  5. Acknowledgment
  6. REFERENCES
  7. Supporting Information

The BLM approach is an excellent example of integrating mechanistic understanding from multiple disciplines into a practical approach to improve application of ecotoxicological data to risk assessments and regulations. However, BLM efforts also have illustrated various problems that can befall such modeling efforts, presenting challenges for its further development and application.

Several reviews [5, 17, 18, 20, 22] have noted that BLMs do not address various toxicodynamic processes potentially important to toxicity predictions. The original concept of the BLM (and the GSIM before it) focused on bioavailability—how complexing ligands and competing cations affect metal accumulation relative to total metal concentration in the exposure water. For the original BLM efforts [2, 3], and most others to date, the toxicodynamic portion of the model is simply to relate a discrete effects endpoint (e.g., an LC50) to a particular level of accumulation on the biotic ligand (e.g., an LA50). A notable omission is that acute metal toxicity often involves disruption of osmoregulation [5], so that rather than reflecting cation competition, ameliorative effects of calcium might be partly attributable to effects on gill permeability, and ameliorative effects of sodium might simply reflect more favorable ion gradients. Whether and when these toxicodynamic considerations, and others [5, 17, 18, 20, 22], would have enough impact on predictions to justify more complex models has not been established. Nonetheless, such issues need to be taken more seriously in BLM development and warrant more attention.

Other uncertainties concern the bioavailability portions of BLMs. For example, BLMs are typically formulated based on bulk water chemistry, but this can differ markedly from the microenvironment at the organism surface, something that was considered by Playle and associates [13]. The resulting disequilibria create uncertainties, especially for the meaning of cation binding constants to surface sites. The BLM parameters referenced to bulk water chemistry need to be considered conditional on the organism and exposure conditions used for their estimation, making extrapolation to other circumstances uncertain. Another bioavailability uncertainty continues to be metal complexation by dissolved and colloidal organic matter. The degree of such complexation can vary widely with the nature of the organic matter and has been reported to produce significant differences among model formulations [17, 22, 23]. The lack of an adequately validated model for broadly addressing complexation to natural organic matter means that further development efforts are needed.

Another general issue regarding BLM formulation that has been raised [5, 17, 18, 20, 22] regards the effects of time. The chemical speciation and surface binding components of biotic ligands are equilibrium models, and the kinetics of toxicodynamic processes also are not addressed in typical BLM efforts. Again, it is not generally established whether and when these considerations would have enough impact on predictions to justify more complex models. However, the kinetics of complexation of metals by humic matter can drastically affect interpretation of toxicity tests when insufficient equilibration time is allowed for the exposure solutions, which was well demonstrated in another “Top 100” paper, by Ma et al. [4]. This study underscored the importance of toxicity tests controlling and documenting exposure conditions, which is needed for effective development of BLMs but has often been inadequate.

Even if these structural model issues are not of significant concern (i.e., the simple BLM formulations are adequate), parameterization of BLMs faces a variety of challenges and uncertainties. Binding constants of the toxic metal and competing cations to the biotic ligand have been estimated in a variety of ways, based on how exposure conditions affect either metal accumulation in gill tissue or whole organisms, metal flux rates, or toxic effect concentrations, each approach having weaknesses [22]. For example, total metal accumulation measurements are just surrogates for the actual biotic ligand, which is a small subset of any total metal measurement, necessitating an assumption of similar binding across diverse sites. Binding constants inferred from water-based effects concentrations arguably reflect the toxicologically active sites but require assuming the model structure is correct and complete. These and other parameter uncertainties [5, 17, 18, 20, 22], and their consequence to model predictions and interpretation, need to be better addressed in BLM development and application.

A final caution regards ascribing undue mechanistic significance to parameter values and model behavior when BLMs are calibrated to water-based effects concentrations. One example of this is inferring an accumulation-based effects concentration such as an LA50, which is inseparable from the binding constant for the toxic metal and the binding capacity, so requires assigning values to these other parameters. The uncertainty of the values for these other parameters, especially if extrapolated from other organisms, as well as structural model uncertainty, will usually make such LA50 estimates uninformative and not useful except as a reference point for model calculations [3]. Efforts would be better served by replacing the LA50, binding capacity, and toxic metal binding constant with an LC50 at some reference conditions for the model parameterization, especially for comparing the sensitivity of different organisms.

An even more troublesome example is the understandable but inappropriate inclination to address lack of model fit by adding or adjusting additional model parameters, especially designating additional metal species to be toxic. As discussed, lack of model fit can be attributable to a variety of factors. To arbitrarily ascribe it to an error in a particular parameter or to the toxicity of additional species requires definitive evidence, from either suitable information independent of the water-based effects concentrations or some diagnostic feature of those concentrations. To do otherwise weakens the mechanistic justification for the relationships and can lead to questionable beliefs about mechanisms.

The BLM approach arose because of recognition of some mechanisms important to the observed variation of metal toxicity across exposure conditions and can provide considerable value to aquatic risk assessments. Its effectiveness and success, and its regulatory relevance, benefit from this mechanistic basis, which will suffer if BLM development becomes too much a data-fitting exercise based on faith that simple implementations of the approach are sufficient. More care is needed in further BLM development and application so that its mechanistic roots are maintained and improved.

Acknowledgment

  1. Top of page
  2. GENESIS OF THE BLM APPROACH
  3. CHALLENGES FOR THE BLM APPROACH
  4. SUPPLEMENTAL DATA
  5. Acknowledgment
  6. REFERENCES
  7. Supporting Information

This document has been reviewed in accordance with USEPA policy and approved for publication. The views expressed in this paper are those of the author and do not necessarily reflect the views or policies of the USEPA.

REFERENCES

  1. Top of page
  2. GENESIS OF THE BLM APPROACH
  3. CHALLENGES FOR THE BLM APPROACH
  4. SUPPLEMENTAL DATA
  5. Acknowledgment
  6. REFERENCES
  7. Supporting Information
  • 1
    Erickson RJ, Benoit DA, Mattson VR, Nelson HP Jr, Leonard EN. 1996. The effects of water chemistry on the toxicity of copper to fathead minnows. Environ Toxicol Chem 15:181193.
  • 2
    Di Toro DM, Allen HE, Bergman HL, Meyer JS, Paquin PR, Santore RC. 2001. Biotic ligand model of the acute toxicity of metals. I. Technical basis. Environ Toxicol Chem 20:23832396.
  • 3
    Santore RC, Di Toro DM, Paquin PR, Allen HE, Meyer JS. 2001. Biotic ligand model of the acute toxicity of metals. II. Application to acute copper toxicity in freshwater fish and Daphnia. Environ Toxicol Chem 20:23972402.
  • 4
    Ma H, Kim SD, Cha DK, Allen HE. 1999. Effect of kinetics of complexation by humic acid on toxicity of copper to Ceriodaphnia dubia. Environ Toxicol Chem 18:828837.
  • 5
    Paquin PR, Gorsuch JW, Apte S, Batley GE, Bowles KC, Campbell PGC, Delos CG, Di Toro DM, Dwyer RL, Galvez F, Gensemer RW, Goss GG, Hostrand C, Janssen CR, McGeer JC, Naddy RB, Playle RC, Santore RC, Schneider U, Stubblefield WA, Wood CM, Wu KB. 2002. The biotic ligand model: A historical overview. Comp Biochem Physiol Part C 133:335.
  • 6
    Sprague JB. 1985. Factors that modify toxicity. In Rand GM, Petrocelli SR, eds, Fundamentals of Aquatic Toxicology. Hemisphere, Washington, DC, USA, pp 124163.
  • 7
    Hunt DTE. 1987. Trace metal speciation and toxicity to aquatic organisms - a review. TR 247. Water Research Centre, Marlow, Bucks, UK.
  • 8
    Campbell PGC. 1995. Interactions between trace metals and aquatic organisms: A critique of the free-ion activity model. In Tessier A, Turner DR, eds, Metal Speciation and Bioavailability in Aquatic Systems. John Wiley, Chichester, UK, pp 46102.
  • 9
    Zitko V, Carson WG. 1976. A mechanism of the effects of water hardness on the lethality of heavy metals to fish. Chemosphere 5:299303.
  • 10
    Stephan CE, Mount DI, Hanson DJ, Gentile JH, Chapman GA, Brungs WA. 1985. Guidelines for deriving numerical water quality criteria for the protection of aquatic organisms and their uses. PB85-227049. National Technical Information Service. Springfield, VA, USA.
  • 11
    Morel FMM. 1983. Principles of Aquatic Chemistry. Wiley-Interscience, New York, NY, USA.
  • 12
    Pagenkopf GK. 1983. Gill surface interaction model for trace-metal toxicity to fishes: Role of complexation, pH, and water hardness. Environ Sci Technol 17:342347.
  • 13
    Playle RC, Gensemer RW, Dixon DG. 1992. Copper accumulation on gills of fathead minnows: Influence of water hardness, complexation and pH of the gill micro-environment. Environ Toxicol Chem 11:381392.
  • 14
    Playle RC, Dixon DG, Burnison K. 1993. Copper and cadmium binding to fish gills: Estimates of metal-gill stability constants and modelling of metal accumulation. Can J Fish Aquat Sci 50:26782687.
  • 15
    Kramer JR, Allen HE, Davison W, Godtfredsen KL, Meyer JS, Perdue EM, Tipping E, van de Meent D, Westall JC. 1997. Chemical speciation and metal toxicity in surface freshwater. In Bergman HL, Dorward-King EJ, eds, Reassessment of Metals Criteria for Aquatic Life Protection. SETAC Press, Pensacola, FL, USA, pp 5770.
  • 16
    Bergman HL, Dorward-King EJ. 1997. Reassessment of Metals Criteria for Aquatic Life Protection. SETAC Press, Pensacola, FL, USA.
  • 17
    Niyogi S, Wood CM. 2004. Biotic ligand model, a flexible tool for developing site-specific water quality guidelines for metals. Environ Sci Technol 38:61776192.
  • 18
    Erickson RJ, Nichols JW, Cook PM, Ankley GA. 2008. Bioavailability of chemical contaminants in aquatic systems. In DiGiulio RT, Hinton DE, eds, The Toxicology of Fishes. Taylor and Francis, Boca Raton, FL, USA, pp 954.
  • 19
    US Environmental Protection Agency. 2007. Aquatic life ambient freshwater quality criteria - copper. EPA/822/R/07/001. Washington, DC.
  • 20
    de Polo A, Scrimshaw MD. 2012. Challenges for the development of a biotic ligand model predicting copper toxicity in estuaries and seas. Environ Toxicol Chem 31:230238.
  • 21
    Meyer JS. 1999. A mechanistic explanation for the 1n(LC50) vs. 1n(hardness) adjustment equation for metals. Environ Sci Technol 33:908912.
  • 22
    Slaveykova BI, Wilkinson KJ. 2005. Predicting the bioavailability of metals and metal complexes: Critical review of the biotic ligand model. Environ Chem 2:924.
  • 23
    Craven AM, Aiken GR, Ryan JN. 2012. Copper(II) binding by dissolved organic matter: Importance of the copper-to-dissolved organic matter ratio and implications for the biotic ligand model. Environ Sci Technol 46:99489955.

Supporting Information

  1. Top of page
  2. GENESIS OF THE BLM APPROACH
  3. CHALLENGES FOR THE BLM APPROACH
  4. SUPPLEMENTAL DATA
  5. Acknowledgment
  6. REFERENCES
  7. Supporting Information

All Supplemental Data may be found in the online version of this article. See Table S1 for the number of citations and rank of all the “Top 100” papers, which in this essay are references [1–4].

FilenameFormatSizeDescription
etc2222-sm-0001-SupTab-S1.pdf45KTable S1. (49 KB PDF).

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.