Environmental Risk Assessment of Chemical Mixtures in...

Environmental Risk Assessment of Chemical Mixtures in Aquatic Ecosystems: State of the Science & Evidence Gaps

Authors: TH Hutchinson1 & S Dungey2 June 2011 1Cefas Weymouth Laboratory Barrack Road The Nothe Weymouth Dorset DT4 8UB UK

2Environment Agency Chemicals Assessment Unit Howbery Park Wallingford Oxfordshire OX10 8BD UK

Cefas contract report C5288 (Cefas – EA cooperative report) page 2

Cefas Project C5288 in cooperation with the Environment Agency (England & Wales): Environmental Risk Assessment of Chemical Mixtures in Aquatic Ecosystems: State of the Science & Evidence Needs

Sponsor: Department for Environment, Food & Rural Affairs


Table of contents

1 Executive summary ................................................................................................................................... 4

2 Recommendations .................................................................................................................................... 6

3 Introduction ................................................................................................................................................ 6

3.1 General background ........................................................................................................................ 6

3.2 EU overview .................................................................................................................................... 7

3.3 United States overview ................................................................................................................... 9

3.4 South East Asia overview ............................................................................................................. 10

3.5 OECD overview ............................................................................................................................. 10

3.6 WHO overview .............................................................................................................................. 10

4 Chemical mixture case studies .............................................................................................................. 11

4.1 Natural and synthetic oestrogens .................................................................................................. 11

4.2 Plant Protection Products .............................................................................................................. 13

4.3 Dioxins ........................................................................................................................................... 14

4.4 Summary of prospective environmental risk assessment approaches ......................................... 15

4.5 Summary of retrospective environmental risk assessment approaches....................................... 17

4.5.1 Summary of eco-epidemiology work involving mixtures ...................................................... 18

4.5.2 Environmental exposure monitoring .................................................................................... 20

4.5.3 Whole Effluent Assessment (Direct Toxicity Assessment) .................................................. 21

5 Conclusions ............................................................................................................................................. 23

6 Acknowledgements ................................................................................................................................. 24

7 References ............................................................................................................................................... 24

Annex 1: Chemicals monitored in UK surface waters ............................................................................... 28


1 Executive summary

1.1 There has recently been a marked increase in international scientific and regulatory activity on the risk assessment of chemical mixtures (often termed ‘cumulative risk assessment’), in part reflecting the growing knowledge base on mixture toxicology related to both human health and the environment. This report reviews the current scientific evidence and knowledge gaps pertinent to the environmental risk assessment (ERA) of exposure to chemical mixtures specifically in the context of aquatic ecosystems. While the underlying principles of mixtures toxicology are shared between mammalian and non-mammalian species, this report does not address chemical mixtures with respect to human health (eg air quality or drinking water sources) nor does the report address terrestrial risk assessment issues (eg contaminated land). Additionally, this report focuses only on chemical mixtures and does not address interactions of chemicals with other environmental factors that are known to affect their impacts on aquatic organisms (eg temperature). 1.2 As noted by the IGHRC (2009) and the ACHS (2010), while the risks from chemicals are largely currently assessed on the basis of exposure to and toxicity from single compounds, humans and wildlife are typically exposed simultaneously to large numbers of chemicals. The IGHRC acknowledged the fact that current risk assessment practices do not generally take chemical mixtures into account. The IGHRC report (2009) recommended a case-by-case strategic approach to the assessment of mixtures based on the best available science, and offered a general framework designed to help risk assessors think about how to address the risks of mixtures to human health, through the stepwise identification of key issues to be considered depending on the type of mixture assessed and the types of data available. While the ACHS regarded the IGHRC (2009) report as being a well thought out approach to human exposures of chemical mixtures, the ACHS concluded that the IGHRC approach would require significant adaptation to meet the needs of environmental risk assessment for multiple species. 1.3 In terms of addressing the ecosystem impacts of real world mixtures, recent progress has been made in terms of eco-epidemiology studies in the UK and in other countries. This retrospective risk assessment approach to real world chemical mixtures is particularly relevant to the Water Framework Directive and the Marine Strategy Framework Directive. The key issue is whether ecotoxicological standards are sufficiently protective for mixture effects, and the different studies only provide partial insight on this point. For example, predicted effects of chemical mixtures (referred to as “toxic pressures”) were found to be positively correlated with changes in fish species assemblages in US rivers (Posthuma and De Zwart 2006). Similarly, an analysis of UK data by the same researchers found highly significant associations between acute toxic pressure arising from mixture exposure and abundance data for more than half the freshwater organism families monitored (EA 2008). These two studies suggest that freshwater species loss at a local scale might be attributed in part to exposure to chemical mixtures, but they were not designed to give any indication of the protectiveness or otherwise of ecotoxicological standards. In contrast, other studies suggest that environmental quality standards are protective for benthic macroinvertebrate family richness (eg Crane et al. 2007). It is important to note that these studies are a product of the limited data available (with their inherent limitations), and correlation is not the same as causation. 1.4 In terms of retrospective risk assessment of real world mixtures, the UK and a number of other countries have successfully used biological effects based approaches (eg Whole Effluent Assessment and Direct Toxicity Assessment) for several years to address complex chemical mixtures in aquatic ecosystems. Together with eco-epidemiology studies, such biological effects based approaches to complex environmental mixtures can often provide highly useful information to support policy needs (eg Water Framework Directive and Marine Strategy Framework Directive). Complementary to these approaches, it is also important to consider the toxicological


mode of action of chemicals known or suspected to be present in aquatic environments where biological effects have been observed. The observations of intersex fish and the measurement of oestrogenic chemicals in UK rivers is an important example of the successful application of the biological effects based approach (Jobling et al. 1998). 1.5 In terms of prospective risk assessments of chemical mixtures, the European Commission has noted that current regulatory approaches to the assessment of chemicals are usually based on the evaluation of single substances, whereas there are concerns that this does not provide sufficient protection from combination effects. They therefore propose that the combination effects of chemicals should be addressed in a systematic way. Acknowledging these concerns, the Council of Environment Ministers adopted certain conclusions on the combination effects of chemicals in December 2009 and invited the Commission to assess how and whether existing legislation addresses this problem and to suggest appropriate modifications and guidelines. DG Environment had also contracted an important study from 2007 to review the current scientific knowledge and regulatory approaches (termed the "State of the Art Report on Mixture Toxicity" by Kortenkamp et al. 2009). 1.6 Currently, the EU Scientific Committee on Health and Environmental Risks (SCHER) has been asked by the Commission to address a number of key questions relating to both the human and environmental risk assessment of chemical mixtures and to provide its final opinion to the Commission by June 2011 at the latest. Key issues of importance for the ERA of chemicals put to SCHER by the Commission include: an assessment of the scientific evidence that when organisms are exposed to a number of different chemical substances, that these substances may act jointly in a way that affects the overall level of toxicity (eg addition, antagonism or synergy); if different chemical substances to which the environment is exposed can be expected to act jointly in a way which affects their impact on the environment, whether or not current assessment methods take proper account of these joint actions; consider possible default approaches for assessing chemical mixtures; and identify the major knowledge gaps with regard to the systematic assessment of the toxicity of chemical mixtures in the context of EU legislation. 1.7 With respect to the current scientific evidence from in vitro and in vivo studies relevant to aquatic organisms, significant progress has been made in recent years towards understanding the mode-of-action of chemicals that underpin whether they may act jointly in a way that affects the overall level of toxicity (eg addition, antagonism or synergy, etc). For example, many studies support the use of the concentration addition approach for chemicals of a similar mode-of action (eg oestrogens) in a given species but it should be noted that this approach needs to consider comparable exposure concentrations and durations in different aquatic species. Current scientific evidence suggests that synergy between chemicals can occur in certain situations and is also best addressed using a toxicological mode-of-action approach (eg chemical synergists used to enhance the effectiveness of insecticides). 1.8 Historically default approaches such as assessment (uncertainty) factors have been widely used in ERA to extrapolate laboratory ecotoxicity data on single chemicals to the real world. For example, the OECD (1995) recommended an assessment factor of 1000 to be applied to the most sensitive acute data for at least three aquatic species, but noted that this was “not based on any theoretical model but has developed in line with experience in effects assessment”. It is unclear to what extent current assessment factors are intended to account for chemical mixture effects (eg ECHA (2008) does not explicitly refer to mixture effects as a contributing factor to assessment factors, although they were mentioned in earlier guidance (eg EC 2003)). With respect to REACH, a recent Swedish study suggested two principal options for considering chemical mixtures, namely a default mixture assessment factor (MAF) and (b) a scenario specific cumulative risk assessment (CRA) (Backhaus et al., 2010).


2 Recommendations

2.1 Consideration should be given to conducting a critical review of the strengths and weakness of assessment (uncertainty) factors in the ERA of laboratory data for single chemicals to chemical mixtures in field situations, taking into account the evidence regarding the toxic mode(s)-of-action of chemicals under acute and chronic exposure scenarios. 2.2 In terms of the Water Framework Directive and Marine Strategy Framework Directive, it could be valuable to consider the potential benefits of further eco-epidemiology studies using freshwater and marine monitoring data to better understand the potential impacts of ‘real world’ chemical mixtures on biodiversity in UK ecosystems. To get the most out of this approach, it might be helpful to perform additional ecotoxicity studies (eg Whole Effluent Assessment techniques) using samples collected from suitable sites, and also assess the number of substances actually present.

3 Introduction

3.1 General background

As observed by the UK’s Interdepartmental Group on Health Risks from Chemicals (IGHRC 2009), current risk assessment practices are largely based on evaluating the toxicity of single chemicals. However, humans are simultaneously exposed on a daily basis to a large number of chemicals, both intentionally and unintentionally. There are regular expressions of media and/or public concern that exposure to this “chemical cocktail” could result in adverse health effects unforeseen by current risk assessment practices. Although not included in the scope of the IGHRC report, similar concerns are also being increasingly raised regarding chemical mixtures present in freshwater and marine ecosystems (Vighi et al. 2003; OSPAR 2005; Chevre et al. 2006; Laetz et al. 2008; ACHS 2010). The IGHRC also recommended that chemical mixtures are best considered as a series of discrete, precisely defined problems for which clear boundaries can be set. Each discrete, precisely defined risk assessment can then be compared to other, similar risk assessments to enable the larger picture to be assembled over time. As with environmental risk assessments, a key factor in risk assessments for chemical mixtures and human health is the availability, or absence, of reliable data for the whole mixture or its components (IGHRC 2009). However, the Advisory Committee on Hazardous Substances noted that if the IGHRC approach were to be developed for environmental assessments, adaptations would have to be made in some generic approaches and to specific areas covering the areas of environmental fate and behaviour, exposure and effects1. It is beyond the scope of this report to review the immense body of published work on chemical mixtures (the first reported experimental studies date back to pre-1900). Scientific evidence of the toxicological importance of chemical mixtures is certainly not new and many of the concepts of mixture toxicology pertinent to environmental species have their origins in medical toxicology research. For example, Bliss (1939) described three types of joint action that chemicals in mixtures may show, namely: (1) independent joint action, where chemicals act independently of one another and have different modes of toxic action; (2) similar joint action, where chemicals have similar effects but do not interact (i.e. dose addition); and (3) synergistic/antagonistic action, where the toxicity of a mixture may be greater than/less than the toxicity that would be predicted from the individual constituents. These concepts are used today by scientists to categorise the behaviour of chemicals in mixtures, in the context of both mammalian and non-mammalian toxicology. For

1 http://archive.defra.gov.uk/environment/quality/chemicals/achs/documents/achs-ighrc-chemicals-framework.pdf


further information on the principles of chemical mixture toxicology see the key reviews by the US EPA (2000), IGHRC (2009), Kortenkamp et al. (2009) and Backhaus et al. (2010). This report summarizes the state of the science for the Environmental Risk Assessment (ERA) of chemical mixtures specifically in the context of aquatic ecosystems. A number of important case studies are included to illustrate the multiple aspects of the mixtures issue. While the basic principles of mixtures toxicology are shared between mammalian and non-mammalian species, this report does not address chemical mixtures with respect to human health (eg airborne contaminants, drinking water sources, etc) nor does the report address terrestrial risk assessment issues (eg contaminated land, sewage sludge disposal to farmland, etc). Additionally, this report focuses only on chemical mixtures and does not address interactions of chemicals with other environmental factors (eg temperature) that are known to affect the health of aquatic organisms (Brian et al. 2008).

3.2 EU overview

Recognizing public concerns over chemical mixtures, the Council of Environment Ministers adopted conclusions on the combination effects of chemicals on 22 December 2009 (see http://register.consilium.europa.eu/pdf/en/09/st17/st17820.en09.pdf). In its conclusions, Council invited the Commission to assess how and whether existing legislation addresses this problem and to suggest appropriate modifications and guidelines. While methodologies for assessing the combination effects of chemicals are being developed and used by scientists and regulators in specific circumstances, as yet there is no systematic, comprehensive and integrated approach. In 2007, DG Environment contracted a study to review the current scientific knowledge and regulatory approaches which led to the in-depth study entitled "State of the Art Report on Mixture Toxicity" (Kortenkamp et al. 2009). For further reference, the state-of-the-art report includes the following sections: Part 1 - The state of the art of mixture toxicology – a critical appraisal of published scientific literature; Part 2 – Identification and description of current provisions from taking into account hazards and risks arising from mixture toxicity in 21 pieces of EU legislation; Part 3 – Survey on approaches and practical experiences in assessing the mixture toxicity of complex environmental samples and waste samples in EU Member States; and Part 4 – Overview of approaches to hazard and risk assessment of chemicals mixtures in the USA, Japan and international bodies.

The EU Scientific Committee on Health and Environmental Risks (SCHER) has been asked by the Commission to address a number of key questions relating to both the human and ERA of chemical mixtures and to provide its final opinion to the Commission by June 2011 at the latest. In addition to human health aspects, key issues of importance for the environmental risk assessment of chemicals put to SCHER by the Commission include:

(i) an assessment of the scientific evidence that when organisms are exposed to a number of different chemical substances, that these substances may act jointly in a way that affects the overall level of toxicity (eg addition, antagonism or synergy, etc),

(ii) if different chemical substances to which the environment is exposed can be expected to act jointly in a way which affects their impact on the environment,

(iii) whether or not current assessment methods take proper account of these joint actions, (iv) possible default approaches for assessing chemical mixtures, (v) the major knowledge gaps with regard to the systematic assessment of the toxicity of

chemical mixtures in the context of EU legislation.

For example, a report for the Swedish National Chemicals Inspectorate (Backhaus et al., 2010) proposes that for prospective assessments in regulatory contexts (eg REACH) the classic toxicological concepts of concentration addition and independent action seem to be the most promising methods. They state that especially concentration addition has proven to provide


generally good approximations of expected mixture toxicities (at least for short-term acute effects) for a wide range of mixtures, exposed organisms and biological endpoints. They also suggest that concentration addition also allows the expected mixture toxicity (EC50 values) to be predicted using the toxicological and ecotoxicological data that may be available from the registration of a compound. In terms of prospective risk assessments under REACH, Backhaus et al. (2010) also concluded that registration of a compound within REACH would allow two principal options for considering chemical mixtures: (a) a default mixture assessment factor and (b) a scenario specific cumulative risk assessment (CRA). In theory, the flexible use of these two options suggested by Backhaus et al. (2010) should be applicable to both acute and chronic aquatic toxicity data. The additional assessment factor idea is commented on further in Section 4.4.

For Plant Protection Products, EFSA is working on cumulative risk assessment which aims to develop methodologies to assess the cumulative effects resulting from consumer exposure to pesticides (see http://www.efsa.europa.eu/en/prapertopics/topic/pesticides.htm). As part of EFSA’s broader work on cumulative risk assessment, in 2006 a “Scientific Colloquium on Cumulative Risk Assessment” was held which helped guide further developments in the field (see http://www.efsa.europa.eu/en/colloquiareports/colloquiapesticides.htm). In 2008, the Plant Protection Products and their Residues (PPR) Panel2 issued an opinion on all types of combined toxicity of pesticides, including the interaction of different chemicals. It concluded that only cumulative effects from concurrent exposure to substances which have a common mode of action raised concerns and needed further consideration. EFSA’s work looks at those groups of pesticides which have similar chemical structure and toxic effects to see if their impact on human health should be assessed collectively rather than just on an individual basis. To our knowledge, no environmental risk assessment is required at the current time.

The European Parliament and Council Directive 98/8/EC for biocidal products contains a stipulation that the risk assessment of a biocidal product should combine the results for the active substance together with the results for any substance of concern to produce an overall assessment for the biocidal product itself, taking account of any likely synergistic effects. For biocidal products containing more than one active substance any adverse effects shall also be combined to produce an overall effect for the biocidal product itself. The proposed new EU Biocidal Products Regulation (5604/1/11 ENV, REV 1 of 2 March 2011) goes further. It states that where the evaluating competent authority considers that there are concerns with regard to the cumulative effects from the use of biocidal products containing the same active substance, it shall document its concerns as part of its conclusions. The evaluation of a biocidal product shall also take into account cumulative or synergistic effects associated with the relevant individual components of the biocidal product, as well as the use of biocidal products containing the same active substance(s)..

In terms of veterinary medicinal products (Directive 2001/82/EC) and human medicinal products (Directive 2001/83/EC) an environmental risk assessment is required for all new marketing authorisation applications for a medicinal product through a centralised, mutual recognition, decentralised or national procedure. Based on currently available information there is currently no requirement to consider mixtures in such an ERA.

With regard to retrospective risk assessment, consideration needs to be given to process-oriented forms of legislation (eg the Integrated Pollution and Prevention Control Directive (IPPC)) and media-oriented European legislation (eg the Water Framework Directive (WFD) and Marine Strategy Framework Directive (MSFD)). It is generally agreed that the Whole Effluent Assessment (WEA) approach provides useful information for complex effluents that contain a large variety of unidentified substances and similar methodologies are used outside Europe to retrospectively address complex chemical mixtures in order to protect aquatic ecosystems (Vighi & Calamari 2 The PPR Panel provides scientific advice on issues that cannot be resolved within the peer review of active substances, or when further scientific guidance is needed on more generic issues in both human and environmental risk assessment of pesticides


1996; Vighi et al. 2003; Whitehouse et al. 2004; Syberg et al., 2009). For example, the OSPAR (2005) WEA approach is used to support the objectives of the OSPAR Hazardous Substances Strategy. WEA consists of a variety of (biological) tests to determine persistence, bioaccumulation and toxicity (PBT-criteria). These are the same criteria that are used within OSPAR’s Hazardous Substances Strategy, with the difference that WEA tests are applied to the entire effluent sample instead of to the individual substances. This means that not only the effects of the known substances, but also those of unknown substances are measured. WEA therefore allows identification of adverse effects that result from substances that cannot be chemically identified or substances that have been identified but have not yet been assessed for their PBT properties. Furthermore, WEA measures the overall combined effects of constituent chemicals. It is noted that the UK government’s report Charting Progress 2: The State of the UK Seas highlighted the importance of addressing the knowledge gaps around biological impacts of chemical mixtures (Defra 2010).

3.3 United States overview

In terms of retrospective risk assessment, since 1972 under the Clean Water Act (also known as the Federal Water Pollution Control Act), the US EPA has implemented the National Pollutant Discharge Elimination System (NPDES) which controls water pollution by regulating point sources that discharge pollutants into waters. Similar to the European Whole Effluent Assessment approach using biological effects assessment for complex chemical mixtures, the NPDES system include Whole Effluent Toxicity (WET) testing using freshwater or marine organisms of direct relevance to the geographic area of interest (Grothe et al. 1996). With regard to retrospective risk assessment for persistent organic pollutants (POPs) such as dioxins, United States actively considers mixtures through a Toxicity Equivalency Factors (TEFs) approach which can be applied to aquatic organisms (Wan et al. 2010)3. The term "dioxin" is commonly used to refer to a family of toxic chemicals that share a similar chemical structure and induce harm through a similar mechanism. Examples of dioxins include polychlorinated biphenyls (PCBs), polychlorinated dibenzodioxins (PCDDs), and polychlorinated dibenzofurans (PCDFs). PCBs were produced commercially in large quantities until production was stopped in 1977. PCDDs and PCDFs are not commercial chemical products, but are unintentional by-products of most forms of combustion and several industrial chemical processes. Dioxins have been characterized by the US EPA as likely human carcinogens and are anticipated to increase the risk of cancer at background levels of exposure. In the UK, the Environment Agency is responsible for monitoring dioxin releases under the IPPC Directive using, for example, the TEF guidance from Defra (2002) and other relevant sources of information. In terms of prospective risk assessment of chemicals such as biocides, pharmaceuticals or plant protection products, to our knowledge there are no specific legal requirements in the US for ERAs to be conducted for chemical mixtures other than using assessment factors as per the OECD Aquatic Guidance Document (OECD 1995). Importantly, however, since 1986 the US EPA has developed a framework for conducting cumulative risk assessments for human health. In assessing cumulative risks, EPA evaluates the potential for people to be exposed to more than one pesticide at a time from a group with an identified common mechanism of toxicity. These cumulative assessments consider exposures from food, drinking water, and residential sources. The US EPA also incorporates regional exposures from residential and drinking water sources since this is the most appropriate way to account for the considerable variation in potential exposures across the country. EPA's cumulative assessments, therefore, approximate as closely as possible people's actual exposures and potential risks resulting from current uses of these pesticides in different parts of the country (see also http://www.epa.gov/pesticides/cumulative/ ). So far, the US EPA has prioritized four groups of pesticides which have a common mechanism of

3 For more information dioxin TEFs as a central tool in protecting human health and for conducting ERAs see http://cfpub.epa.gov/ncea/CFM/nceaQFind.cfm?keyword=Dioxin


toxicity and require cumulative risk assessments because exposure to these pesticide groups may pose potential risks to human health and the environment. The four groups are: the organophosphates, N-methyl carbamates, triazines and chloroacetanilides. While not specific to environmental risk assessment, the principles developed by the US EPA are being considered through WHO and OECD activities and could indirectly influence future environmental risk assessment requirements. For example, the US EPA use of the chemical’s mode of action and mechanism of toxicity to underpin the cumulative risk assessment, and the mode of action approach is also being considered to support the intelligent ecotoxicology testing in Europe (ECETOC 2007).

3.4 South East Asia overview

To our knowledge, there are currently no policy developments for addressing the environmental risks of mixtures of agrochemical, biocidal or industrial chemicals in Japan or other south east Asian countries. However, Japanese scientists and regulators are actively involved in the cumulative risk assessment initiatives led by the WHO and OECD regarding persistent organic pollutants using the Toxic Equivalent (TEQ) and similar approaches (Kajiwara et al. 2007).

3.5 OECD overview

The OECD is working in close cooperation with the WHO and other organisations under the umbrella of the Inter-Organization Programme for the Sound Management of Chemicals (IOMC; see http://www.who.int/iomc/en/) (see the overview of WHO activities on combined exposures to multiple chemicals for more information). In addition, the OECD has provided a guidance document on the aquatic toxicity testing of difficult substances and mixtures (eg PCBs and other lipophilic chemical mixtures) as part of the Test Guidelines Programme (OECD 2000). The OECD (1995) guidance document for aquatic effects assessment briefly discussed the use of assessment (uncertainty) factors in the ERA of laboratory data on single chemicals to the real world and recommended, for example, an assessment factor of 1000 applied to the most sensitive acute data for at least three aquatic species but notes “they are not based on any theoretical model but have developed in line with experience in effects assessment”. As discussed at the recent WHO OECD ILSI-HESI mixtures workshop, it is important that when assessment factors are used then they should be transparent and evidence based.

3.6 WHO overview

WHO is actively working on the cumulative risk assessment of combined exposures to multiple chemicals. A WHO/IPCS workshop on cumulative risk assessment was held in the Washington DC in March 2007 and was focussed on developing a framework for the human health risk assessment of combined exposures to multiple chemicals. More recently, the February 2011 workshop of the WHO OECD ILSI/HESI in Paris included discussion of the WHO cumulative risk assessment framework (Figure 1) from the perspective of both human health and ERA. Since the basic elements of this promising framework are built around the principles of exposure assessment and hazard assessment, taking into account a mode of action approach, there is significant potential to apply or adapt the WHO framework for application to ERAs. Recommendations made at the WHO OECD ILSI/HESI workshop included conducting an environmental case study of the WHO framework (eg using natural and synthetic oestrogens).


Figure 1. WHO framework for the cumulative risk assessment of chemical mixtures. See: http://www.who.int/ipcs/methods/harmonization/areas/aggregate/en/index.html. Notes: The point of departure (POD) is a selected measure of effect. It may be a no- or lowest-observed-(adverse-)effect level (NO(A)EL or LO(A)EL) or a dose or concentration associated with a specified increase in the incidence of an effect (e.g. a benchmark dose or concentration associated with a 10% increase in incidence of an effect). In the WHO framework, the mode of action (MOA) is a biologically plausible sequence of key events leading to an observed effect, supported by robust experimental observations and mechanistic data. The margin of exposure (MOE) is the ratio of the selected measure of effect to the estimated exposure dose or concentration. Components of the framework are described in the WHO consultation document (WHO 2009).

4 Chemical mixture case studies

4.1 Natural and synthetic oestrogens

Over the past decade, a large number of natural and synthetic chemicals have been identified as endocrine active, due to their ability to mimic endogenous hormones. Endocrine active chemicals mediate their effects by binding to hormone receptors as agonists or antagonists, by inhibiting the enzymes responsible for steroid hormone biosynthesis, and/or by inducing the enzymes responsible for steroid metabolism. In many countries, including the UK, particular focus has fallen on chemicals which mimic 17β-oestradiol (E2) by binding to the oestrogen receptor(s) (ERs) to elicit agonist and(or) antagonistic responses. These chemicals are collectively described as xenoestrogens. With the exception of the synthetic steroids, the xenoestrogens discovered so far are only weakly active when compared with endogenous steroids. However, evidence for reproductive abnormalities that are indicative of sex hormone disruption in wild fish populations, supported by in situ monitoring studies with caged fish, implies that some aquatic environments


contain xenoestrogens at concentrations high enough to be of concern to wildlife (Jobling et al. 1998). In the UK, natural estrogens such as E2 and oestrone, synthetic steroids including, 17α-ethynylestradiol (EE2), and other nonsteroidal chemicals known to have oestrogenic effects (such as alkylphenols), have been detected in effluents that discharge into rivers. The plausibility of this real world mixture scenario in the UK, together with a large number of in vitro studies on oestrogenic responses of mixtures (Silva et al. 2002; Kortenkamp et al. 2007), led Thorpe et al. (2001 & 2003) to evaluate the in vivo significance of oestrogens to fish. Large-scale laboratory experiments were conducted to assess the in vivo potency of binary mixtures of oestrogenic chemicals using plasma vitellogenin (VTG) concentrations in juvenile rainbow trout (Oncorhynchus mykiss) as the endpoint. The oestrogenic potencies of 17β-estradiol (E2), 4-tert nonylphenol (NP), and methoxychlor (MXC) were determined following 14-day exposures to the individual chemicals and binary mixtures of these chemicals. E2, NP, and MXC all induced concentration-dependent increases in plasma VTG (Thorpe et al. 2001). Concentration-response curves for fixed ratio binary mixtures of E2 and NP (1:1000), E2 and MXC (1:1000), and NP and MXC (1:1) were compared to those obtained for the individual chemicals, using the model of concentration addition. Mixtures of E2 and NP were additive at the concentrations tested, but mixtures of E2 and MXC were less than additive. This suggests that while NP probably acts via the same mechanism as E2 in inducing VTG synthesis, MXC may be acting via a different mechanism. In fact, due to the neurotoxicity of MXC, it was not possible to determine whether mixtures of MXC and NP were additive using VTG induction, because the toxicity of MXC restricted the effect range for which the expected response curve for the binary mixture could be calculated. The work by Thorpe et al. (2001) clearly illustrated that the model of concentration addition can accurately predict effects on VTG induction, where we know that both chemicals act via the same mechanism in mediating a vitellogenic response. Subsequently Thorpe et al. (2003) successfully defined the relative potency estimates for the natural steroids E2 and oestrone (E1) and the synthetic steroid EE2 to fish in vivo. All of these steroids bind to the oestrogen receptor(s) and have been shown to elicit a range of estrogenic responses in fish at environmentally relevant concentrations. In this study the estrogenic activity of E2, E1, and EE2, and the combination effects of a mixture of E2 and EE2 (equipotent fixed-ratio mixture), were assessed using vitellogenin induction in a 14-day in vivo juvenile rainbow trout screening assay. Median effective concentrations, relative to E2, for induction of vitellogenin were determined from the concentration response curves and the relative oestrogenic potencies of each of the test chemicals calculated. Using the model of concentration addition it was shown that this activity of the binary mixture could be predicted from the activity of the individual chemicals. Thorpe et al. (2003) concluded that due to the ability of each individual steroid to contribute to the overall effect of a mixture, even at individual no-effect concentrations, combined with the high oestrogenic potency of the steroids, and emphasized the need to consider the total estrogenic load of these chemicals in UK rivers. More recently, Brian et al. (2005) investigated the combined effects of a multicomponent mixture of five estrogenic chemicals using vitellogenin induction in male fathead minnows as an end point. The mixture consisted of 17β-oestradiol, 17α-ethynylestradiol, nonylphenol, octylphenol, and bisphenol-A. After determining concentration–response curves for each of the chemicals individually, the chemicals were then combined at equipotent concentrations and the mixture tested using fixed-ratio design. The effects of the mixture were compared with those predicted by the model of concentration addition using biomathematical methods, which revealed that there was no deviation between the observed and predicted effects of the mixture. These findings demonstrate that estrogenic chemicals have the capacity to act together in an additive manner and that their combined effects can be accurately predicted by concentration addition. Brian et al. (2005) also explored the potential for mixture effects at low concentrations by exposing the fish to each chemical at one-fifth of its median effective concentration (EC50). Individually, the chemicals did not induce a significant response, although their combined effects were consistent with the predictions of concentration addition. Again, this demonstrates the potential for oestrogenic chemicals to act additively at environmentally relevant concentrations owing to their common MOA


in fish. They concluded that their findings again highlight the potential for ERA procedures to underestimate the hazard posed by mixtures of chemicals that act via a similar mode of action, thereby leading to erroneous conclusions of absence of risk. An essential point in seeking to address real world mixtures is to take exposure into account, an example of which is the work by Sumpter et al. (2006) who used the GREAT-ER exposure model to predict the concentrations and then the oestrogenic effects on fish, of a mixtures of chemicals (including steroidal oestrogens and nonylphenol) both individually throughout an entire UK river catchment and as a mixture. In summary, a series of UK studies conducted over a decade clearly shows that evidence on exposure together with knowledge of the MOA (in this case oestrogenicity) can be used to guide an intelligent testing strategy and environmental risk assessment.

4.2 Plant Protection Products

There are a wide range of economically important pesticide active ingredients used today to support food production and prevent waste during crop storage. It is beyond the scope of this report to review the large number of papers published on pesticide mixtures (for examples see Faust et al. 2001; Backhaus et al. 2004; Chevre et al. 2006; Junghans et al. 2006). In certain farming situations, co- application of pesticides through approved tank mixes may be used. A key study by Matthiessen et al. (1988) examined the joint acute toxicity to rainbow trout of eleven tank-mixes composed of pairs of six fungicides and herbicides (prochloraz, fenpropimorph, diclofop-methyl, tridemorph, benzoylprop-ethyl and propiconazole) using 96 h LC50 tests. The toxicity of the mixtures ranged from being half of that expected on the basis of additive toxicity of the components, to less than 1.4 times the expected value. In view of the degree of experimental error inherent in the LC50 determinations, they concluded that these lethality data provided no evidence for the existence of synergistic (i.e. more-than-additive) toxicity of the respective tank-mixes. In terms of other reports of pesticide synergy, Laetz et al. (2008) studied the potential for mixtures of organophosphate and carbamate pesticides to adversely affect Pacific salmon (Oncorhynchus sp.) given evidence of these insecticides being detected in freshwater ecosystems. Since both these pesticide classes inhibit the activity of acetyl cholinesterase (AChE), they assessed whether such an insecticide mixture produced additivity, antagonism or synergistic toxicity. Brain AChE inhibition was measured in juvenile coho salmon (Oncorhynchus kisutch) exposed to sublethal concentrations of the organophosphates diazinon, malathion, and chlorpyrifos, as well as the carbamates carbaryl and carbofuran. Concentrations of individual chemicals were normalized to their respective median effective concentrations (EC50) and collectively fitted to a nonlinear regression to determine whether toxicological responses to binary mixtures were additive, antagonistic, or synergistic. They reported both addition and synergism, with a greater degree of synergism at higher exposure concentrations. Several combinations of organophosphates were also lethal to salmon at concentrations that were sublethal in single-chemical trials, leading Laetz et al. (2008) to conclude that single-chemical risk assessments are likely to underestimate the impacts of these insecticides on salmon in river systems where mixtures occur.

More recently, Bjergager et al. (2011) reported the occurrence of the synergistic interactions between prochloraz and esfenvalerate in the microcosms and at environmentally realistic concentrations. The driver for this work was previous laboratory experiments showing that azole fungicides enhance the toxicity of pyrethroid insecticides in the freshwater crustacean Daphnia magna. In fact, a MOA approach to these observations would indicate the disruption of invertebrate detoxification pathways by the azole which concomitantly increases the effective exposure to the neurotoxic pyrethroid. This exposure scenario of pesticide mixtures is also similar to the use of piperonyl butoxide as a potent inhibitor of detoxification enzymes in insect pests, thereby making pest species more vulnerable to the pesticide itself. Since it was discovered by the pesticide industry in the 1940’s, piperonyl butoxide has been used successfully to act as a synergist for insecticides such as the pyrethroids and thereby help reduce the amount of insecticide active ingredient applied to crops (Lawler et al. 2008). This brief summary underlines the value of using


the MOA framework for supporting the risk assessment of pesticides and other specifically-acting chemicals, for example as proposed by ECETOC (2007) in Figure 2. Figure 2. ECETOC mode of action (MOA) framework for intelligent ecotoxicity testing (ECETOC, 2007).

4.3 Dioxins

During the last 15 years, WHO through the International Programme on Chemical Safety (IPC) has established and regularly re-evaluated toxic equivalency factors (TEFs) for dioxins and related compounds through expert consultations. WHO-TEF values have been established for humans and mammals, birds and fish. These international consensus TEFs have been developed for application in risk management in various Member States and have been adopted formally by a number of countries and supranational bodies, including, amongst others, Canada, Japan, the United States and the European Union. The Toxic Equivalent (TEQ) method was developed for use with compounds that activate the aryl hydrocarbon receptor (see Defra (2002) for background information). This is a relative potency method that assumes the additivity of doses of individual components of the mixture after normalization of the response to a reference chemical. The Relative Potency Factor method is a generalized form of the TEQ method and has been used for classes of pesticides and other chemicals. This method also uses dose addition as the default assumption for the effects of mixtures.


Table 1. Summary of International Toxic Equivalency Factor Scheme for dioxins and furans (source: Defra 2002).

For further up to date information, see the WHO International Programme on Chemical Safety (IPCS) website http://www.who.int/ipcs/assessment/tef_update/en/.

4.4 Summary of prospective environmental risk assessment approaches

A comprehensive approach to the prospective risk assessment of chemicals is provided by the recently published REACH guidance documents published by the European Chemicals Agency4. The structure of the guidance document is summarized in Figure 3. The output of the dose-response characterisation is a ‘predicted no effect concentration’ (PNEC)5, which is intended to represent a concentration below which an unacceptable effect will most likely not occur in a multi-species ecosystem. They are typically derived by extrapolation from measured (preferably long-term) toxicity data for the most sensitive species in a laboratory environment, using test methods that seek to maximise chemical (bio)availability. A numerical extrapolation is needed to ‘translate’ the laboratory data to the field situation, which takes account of a number of uncertainties (ECHA, 2008b):

• intra- and inter-laboratory variation of toxicity data; • intra- and inter-species variations (biological variance); • short-term to long-term toxicity extrapolation;

4 http://guidance.echa.europa.eu/docs/guidance_document/information_requirements_en.htm 5 PNECs are used for risk assessment under REACH, for biocides and also pharmaceutical active substances. Environmental quality standards for specific pollutants under the Water Framework Directive (to establish good ecological status) are derived in essentially the same way. A slightly different approach is adopted for plant protection products: toxicity data are compared with exposure estimates directly to derive a quotient. This is just a slightly different way of approaching the same issue (the acceptable ‘margin of safety’ is effectively the same as an assessment factor, and these are broadly (though not exactly) the same as for industrial substances).


• single stress (laboratory) to multiple stressor (field impact) extrapolation6. Figure 3. Structure of the chemical safety assessment guidance (ECHA 2008a)

Typically, this extrapolation step uses assessment factors (also known as safety or uncertainty factors), or a combination of assessment factors and modelling approaches. Assessment factors for the aquatic environment in a REACH context are summarised in Table 2. Table 2. Assessment factors used to derive the PNECaquatic (ECHA 2008b) Available data Assessment factor At least one short-term L(E)C50 from each of three trophic levels (fish, invertebrates (preferred Daphnia) and algae)

1000

One long-term EC10 or NOEC (either fish or Daphnia) 100 Two long-term results (eg EC10 or NOECs) from species representing two trophic levels (fish and/or Daphnia and/or algae)

50

Long-term results (eg EC10 or NOECs) from at least three species (normally fish, Daphnia and algae) representing three trophic levels

10

Species sensitivity distribution (SSD) method 5-1 (to be fully justified case by case) Field data or model ecosystems Reviewed on a case by case basis

6 An earlier version of the guidance (EC 2003) included an explicit statement that “additive, synergistic and antagonistic effects from the presence of other substances may also play a role” in relation to this uncertainty. The reason for its omission from the present guidance is not clear, and illustrates the lack of transparency in the derivation of these factors.


The size of the assessment factor depends on the confidence with which a threshold can be derived from the available data, and this is based on precedent and on early surveys of the ratios between acute and chronic toxicity of chemicals (EA, 2003; Chapman et al., 1998). In general, conservative default assessment factors are used, which assume that each of the uncertainties (identified above) makes a significant contribution to the overall uncertainty (ECHA, 2008b). EC (2002) noted that whilst there is substantial evidence to demonstrate the uncertainty described by the first three bullet points for plant protection active substances, there are only a few, if any cases, that support the uncertainty mentioned in the fourth bullet point (i.e. in general, a test organism tends to be more sensitive in a laboratory test than in a mesocosm study). EC (2002) added that the contribution of each of the different factors influencing the overall uncertainty can not easily be quantified and may differ between acute and chronic testing. It is standard practice to derive toxicity thresholds using the default assessment factors described in European technical guidance documents. However, for any given substance there may be evidence that the overall extrapolation uncertainty is low, or that one particular component of the uncertainty is more important than any other. In these circumstances it is permitted to vary (i.e. raise or lower) the size of the assessment factor depending on the available evidence7. ECHA (2008b) also adds that the default assessment factors may be modified under some circumstances, which include:

• evidence from structurally similar compounds; and • knowledge of the mode of action, including endocrine disrupting effects.

Historically, it has therefore generally been accepted that there is an implicit assumption that assessment factors are intended to account for mixture effects, at least to some extent, and that these can be varied on the basis of evidence. Nevertheless, we currently have only a weak understanding of the level of uncertainty that mixture interactions introduce. Backhaus et al. (2010) suggested the use of a default mixture assessment factor for use in PNEC derivation. Whether this is actually needed in all cases is unclear, and the scientific basis would need to be carefully explained (eg should different factors be applied to acute and chronic data sets?). Simply increasing the level of precaution in the PNEC is likely to identify more scenarios as posing a risk. There therefore needs to be a consideration of the contribution of other areas of uncertainty to the assessment factor as well as possibilities for refinement should a risk be identified. In this regard, Chapman et al. (1998) encouraged the use of experimental evidence rather than defaulting to safety (assessment) factors to compensate for lack of information. Although this is a reasonable suggestion, the limits of testing requirements for regulatory purposes and large number of chemicals supplied may restrict the extent to which it can be put into practice. This is an issue that could warrant further review.

4.5 Summary of retrospective environmental risk assessment approaches

From a UK perspective, current Environment Agency data show that rivers in England and Wales are less polluted than they were two decades ago, on the basis of both biological and chemical measures8. Improved ecosystem health is indicated by such phenomena as the return of otters to most lowland river catchments (following extermination from many areas in the 1970s primarily due to pesticide pollution9), and the fact that coarse fish are now found in more waters than at any time

7 In general, the minimum factor of 10 (applied to chronic data sets) cannot be decreased on the basis of laboratory studies alone. However, where data are plentiful, statistical modelling approaches may be used and assessment factors tend to be smaller, reflecting reduced uncertainty. 8 http://www.environment-agency.gov.uk/research/planning/34383.aspx 9 http://www.environment-agency.gov.uk/static/documents/Business/Otters_the_facts.pdf


during the last century10. Nevertheless, populations of some fish species (such as salmon) are still low or have declined rapidly to critical levels (e.g. European eel). Whilst a variety of factors are likely to be involved (including unsustainable fishing practices, parasites/disease, and changes to habitat caused by climate and human factors), pollution is thought to make a contribution3. In general terms, current approaches to chemical risk management are presumed to be effective at protecting the aquatic environment from unacceptable damage due to the presence of specific pollutants. However, we know that chemical contaminants rarely occur in isolation but in combination with other substances and other stressors (e.g. varying flow and sedimentation rates, nutrients, etc.). Indeed, there is some recent evidence from eco-epidemiology studies that the presence of pollutant mixtures may be a significant factor in impacting freshwater biodiversity (see section 4.5.1). This in turn could imply that the current chemical risk assessment approach - based on single chemicals - might not always be sufficiently protective in all circumstances. To understand the impact of mixtures of chemicals on the environment, we need to know their ecotoxicological profiles (and how these change with different component combinations), the substances and their concentrations to which organisms will be exposed, and whether the toxicity thresholds we estimate (below which effects are unlikely to occur) take adequate account of other chemical pressures experienced by aquatic organisms11. One way of investigating this issue is to compare biological quality data against chemical quality data from actual sites. This is known as eco-epidemiology. 4.5.1 Summary of eco-epidemiology work involving mixtures

Posthuma & de Zwart (2006) investigated whether exposure to toxicant mixtures is associated with fish assemblage characteristics in the field, and described the relationships between predicted chronic and acute mixture risks and observed impacts. They compiled and analyzed fish abundance and abiotic monitoring data from surface waters in Ohio, USA. Exposure assessment, risk assessment with species-sensitivity distributions, and mixture toxicity rules were used to calculate a relative risk predictor, called the multi-substance potentially affected fraction of species (msPAF). The substances included in the analysis were five heavy metals, ammonia, triclosan, linear alkylbenzenesulfonate, alcohol ethoxylates, alcohol ethoxylate sulfates and boron. Predicted acute and chronic risks ranged from low values to more than 10 and 50% of species potentially affected, respectively. The results showed that exposure to toxicant mixtures (expressed as msPAF) in field conditions was significantly associated with altered fish assemblage characteristics, when these are expressed as species-specific abundance and occurrence parameters (although there was a large scatter around the regression line between predicted risk and impact assigned to mixture toxicity). The observed loss of species at specific locations is consistent with predicted acute risks from chemicals. For the chronic risk predictor, the maximum observed fraction of lost species was approximately one-tenth of the predictor value. There was no association between predicted risk and ecological summary parameters (e.g. total abundance, and measures of overall species diversity), which was thought to be due to the large number of stressor-response types in the environment. Overall, the study suggests that toxicant mixture effects are important when considering a multiple stress situation. On the other hand, there is some evidence obtained from field data suggesting that the toxicity thresholds we derive from laboratory studies on single species and single chemicals may not be under-precautionary. For example, Crane et al. (2007) analyzed spatially matched measurements

10 http://www.environment-agency.gov.uk/static/documents/Research/Fisheries_sum_ENG.pdf 11 This document focuses on the aquatic environment, since this is the compartment for which most substances have data. Similar considerations are likely to apply to soil.


of benthic macroinvertebrate family richness and dissolved metal (i.e. cadmium, chromium, copper, iron, nickel, lead, and zinc) concentrations in rivers for two sampling periods (Spring and Autumn) in England and Wales from 1995. They found that the proposed environmental quality standards for the individual metals were similar to dissolved metal concentrations in rivers associated with unimpaired benthic invertebrate assemblages. The only exceptions were iron and zinc, for which the standards were substantially below concentrations associated with impaired invertebrate assemblages in the field (presumably due to the relatively large assessment factors that were used in their derivation, making the standard more conservative than necessary). Mixture effects were implicitly accounted for because the biological responses were the result of all stressors affecting each site, not just the contaminants of interest. It therefore suggests that mixture effects were not especially important for the sites that were examined, although benthic invertebrates may not always be the most sensitive species to specific metals. The Environment Agency subsequently adopted two approaches to compare biological measures of aquatic ecosystem health with chemical toxicity data to help diagnose causes of biological impact so they might be able to provide better focus to River Basin Management Plans under the Water Framework Directive (WFD): 1) A scoping study identified the relative importance of different stressors on aquatic ecosystems

(EA, 2008a). Biological impact was assessed by identifying the number of ‘expected’ invertebrate species that were missing from a number of river sites in England and Wales over the period 1996–2004. Impacts were then compared with matched site-specific data on a number of stressors including habitat characteristics, classical water chemistry (oxygen level, etc.), nutrients, and concentrations of six heavy metals, ammonia, nitrite and a number of pesticides. The summary statistic “acute toxic pressure” was calculated using species sensitivity distributions, firstly on an individual compound basis, then per compound group (defined as having the same toxic mode of action), and finally for the whole mixture of compounds and compound subgroups with similar and dissimilar toxic modes of action. The results of this preliminary exercise suggested that there was a significant association between (mixture) acute toxic pressure and decreased family abundance at a local scale.

2) An alternative approach (EA, 2008b) utilised the same data inputs but used a geographic

information system (GIS)-based weights-of-evidence and weighted logistic regression approach to try and identify possible causes of impact. The spatial association between ten stressor variables (including water chemistry, nutrients, and toxicity for metals and pesticides) and macrofauna impairment was determined by statistically combining stressor concentration maps. This resulted in probability maps that highlight locations within the study area that have the greatest predicted risk of biological impairment, as well as estimates of stressor influence and risk values for the entire study area. These maps were then compared with sample sites that were known to have reduced species richness (compared to expectations from control sites). On average, the models successfully predicted 85% of the biologically impaired sites based on the spatial distribution of various stressor concentrations, and accurately predicted 80% of non-impaired sites. Most (86%) of the false positives produced by the model had some degree of impairment present (i.e. fewer observed macrofauna than expected). The Spring season model for the overall study area was most influenced by the chloride variable, which may serve as a proxy for stress associated with surface run-off. Biological oxygen demand (BOD) was the most influential variable in the Autumn season model for the overall study area, likely representing increases in eutrophication as a result of the seasonal influx of organic matter (leaf fall) combined with existing anthropogenic sources of eutrophication. Ammonia was the most influential variable in the agricultural land use models for both seasons. A poor performance of the pesticide toxicity variable may indicate the need for further refinement of the variable. Acidity (pH) and biological demand (BOD) were important factors in the urban land use model in both seasons. Metal toxicity was most significant in urban land use, particularly during the autumn season. Nitrate, most likely from run-off after crop harvest, was a significant factor in the autumn season for both agricultural and urban land use.


The two studies (EA, 2008a and 2008b) approach the identification of stressors identification and influence in different ways but both seek to identify the main pressures in the presence of multiple stresses on biological communities. The two methods strongly agreed on the causes of impact (75-80% agreement), which is probably not unexpected as they were based on essentially the same data set. However, the agreement in the influence (ranked importance) of identified stressors at sites was much lower, though still significant. Although these were only scoping studies using limited data inputs, it was suggested that both techniques could be used as screening tools for stressor prioritisation in watershed management. A further Environment Agency project (EA, 2011a) built on the earlier field study by Crane et al. (2007). By combining biological and chemical monitoring data from the same locations, it was possible to investigate the relationship between annual average chemical concentrations and biological quality (for data collected between 2006 and 2008) and to distinguish between the impact of a single chemical and the effects of complex mixtures. Forty determinands (including metals, pesticides and sanitary determinands) were included in the analysis. ‘Limiting functions’ (effectively field-based thresholds) were estimated by quantile regression of the biological data plotted against measured chemical concentrations. This is useful for looking at the effects of contaminants when they occur in the presence of potentially confounding factors (e.g. co-exposure to other chemical stressors in a mixture), although it might not be appropriate for cases where potential explanatory variables co-vary in a linear way. Thresholds of chemical exposure consistent with definitions of ‘high’, ‘good’ and ‘moderate’ ecological quality under the WFD were derived for some, but not all, of the chemicals (mainly because of dataset limitations in terms of number of measurements and the high prevalence of monitoring data below the limit of analytical detection). An analysis of co-variance was also undertaken. Statistically significant thresholds of reduced ecological quality in response to stressors were identified for all of the sanitary determinands, but for only a limited number of toxic substances (i.e. aluminium, copper, iron, manganese, mercury and nickel). These thresholds are useful for calibrating the relative stringency of environmental quality standards (EQS) compared with ecological protection goals. In most cases, it was found that proposed EQSs appear to offer adequate protection for ecological communities. These examples show that data can be collected and analysed from the field to assess the impacts of individual chemical stressors when they occur alongside other stressors. However, the analyses are sophisticated and data inputs are technically highly demanding. In considering the results of such comparison exercises, it should be noted that: • correlation is not the same as causation; • the analyses only consider a relatively small number of chemical contaminants; • toxic pressure is often expressed in terms of acute lethality rather than long-term sub-

lethal impacts (the latter are the basis for most toxicity thresholds); • the measurement of biological impairment is often based on qualitative information on

species richness and limited information on abundance, with limited replication. Some taxonomic groups might not be considered at all.

It is therefore important to avoid over-interpretation of the recent UK findings (EA 2008a). 4.5.2 Environmental exposure monitoring

Inputs to the environment are invariably mixtures of chemicals, but these are rarely fully characterised. Exposure is an important factor to consider in the assessment of mixture interactions, but the concentrations of different components will vary with time, environmental compartment and locality.


For commercially produced substances, we generally need at least some information on emission patterns, degradation and partitioning behaviour to be able to predict likely concentrations in different environmental compartments using models. These predictions tend to be conservative, and are often relatively crude (e.g. by assuming a fixed dilution factor for an effluent discharge in the receiving water course). Simply adding predicted concentrations of different components together may be misleading, because the chemicals may not be used in the same localities, or because worst case assumptions will be compounded. More sophisticated probabilistic assessment may be possible, but this is not routinely done at the moment. The complexity will also increase with the number of components that need to be considered, and may be unmanageable for more than a handful of substances at any one time. Monitoring data can provide a more realistic picture of exposure, although there are a number of drawbacks. First, the ability to detect spatial and temporal variation in concentrations depends on the sampling strategy and analytical detection limits. Secondly, most national schemes and published monitoring studies focus on a limited number of chemicals (often heavy metals, and a sub-set of pesticides and other organic substances such as PAHs or oestrogenic compounds). Thirdly, the origin of detected chemicals can be difficult to establish (and in some cases they might be the result of accidental discharge or incorrect disposal). The Environment Agency has developed a rapid screening method using gas chromatography-mass spectroscopy for water and sediment samples to establish semi-quantitative concentrations for over 850 trace organic substances (see Appendix 1). The number of additional substances that may be present in a sample but excluded from the screen (because they are not present in the target database) is unknown. 4.5.3 Whole Effluent Assessment (Direct Toxicity Assessment)

In the UK, Direct Toxicity Assessment (DTA) was developed as a complementary approach to traditional chemical-specific analysis of effluents. It is a broad spectrum technique normally used for defining the short term (acute) toxicity of an effluent discharge using standardised aquatic toxicity tests on algae and invertebrates (in the US for example, the focus is more on sublethal and chronic effects (Grothe et al. 1996)). The toxic response takes account of chemical and ecotoxicological interactions within the effluent, whilst avoiding the need for environmental data on individual substances and expensive chemical analysis. Within the UK, DTA is used to provide a robust prediction of the hazard posed by complex effluents within a risk management framework contributing to integrated water quality management (Figure 4). The quality of surface waters and the control of discharges to those surface waters are governed by a range of national and international (EU) legislative measures. Whilst the aims of much of this legislation are expressed in terms of ‘harm’ to the environment, their practical implementation is generally based on a limited range of individual chemical-specific control requirements, ie the control and measurement of individual chemicals. However, these are not always adequate due to the often large number of chemicals present in discharges and possible interactions between these substances. Bioassays for assessing water quality and controlling emissions to surface waters can help meet the aims of legislation by addressing some of the shortcomings of chemical-specific approaches. DTA can help fill the gaps and be used in a complementary way alongside chemical-specific measures, especially with respect to complex chemical mixtures. The aim is to ensure the survival of aquatic organisms by eliminating acutely toxic effects from point source discharges beyond an agreed point of protection. DTA techniques can be also used to proactively demonstrate low (or no) effluent toxicity or to show reductions in toxicity following risk management activity (Whitehouse et al. (2004) and also http://www.environment-agency.gov.uk/business/regulation/38783.aspx).


Figure 4. The role of Direct Toxicity Assessment in the context of integrated water quality management (Whitehouse et al. 2004).

Whilst DTA can provide an indication of combination effects arising from a complex mixture, it is not possible to identify the substances that are responsible without further investigation. Depending on the complexity of the mixture, this can be difficult and costly. There are also limitations to the usefulness of the data:

a) DTA is normally performed on industrial effluents. Unless the components all behave similarly, it is likely that the composition (and hence mixture effects) will be substantially different once the mixture is released into the environment, due to degradation and partitioning processes, particularly over long timescales. It may therefore have limited relevance.

b) Unless the composition of the mixture is analytically measured during the course of the toxicity tests, it will not be possible to establish whether concentrations of individual components are maintained. This is an important factor because changing concentrations make it very difficult to establish a dose-response relationship. Testing


is not useful if the concentrations of the components vary significantly over the duration of the tests.

c) Combination effects may differ when substances are adsorbed to solid phases (e.g. soil and sediment, with differing organic carbon contents, etc.), due to changes in bioavailability.

Any decision to investigate specific combinations of chemicals therefore needs to clearly take account of the use that the information will be put to. The same is true even for relatively simple mixtures in a laboratory environment. Indeed, guidance for the CLP Regulation emphasises that hazard classification is best done on the basis of the known properties of the components rather than the mixture itself. Animal welfare considerations also limit the possibilities for investigating how changes in mixture composition may affect fish responses.

5 Conclusions

Current prospective risk assessment approaches to estimate PNEC and EQS values are generally considered to be robust, although explicit evidence for including chemical mixture considerations in these estimations is lacking (ECHA 2008b). Importantly, however, prospective risk assessment approaches for the aquatic environment are supported by holistic biological monitoring approaches (Hering et al. 2010) and by whole effluent assessment (Whitehouse et al., 2004; OSPAR 2005). Some eco-epidemiology studies suggest that the presence of certain chemical mixtures in the environment may correlate with reduced biodiversity, whilst others suggest that EQSs do in fact match concentrations associated with sites that have good ecological quality (i.e. that they are protective). These analyses are based on a very small number of substances (and limited biological data), so the wider applicability of their conclusions is unknown. There is therefore scope for additional research, for instance, using the eco-epidemiology approach described by Posthuma and de Zwaart (2006). For example, a much more focussed and detailed eco-epidemiology study could possibly provide further useful insights (e.g. by investigating the appropriateness of existing PNECs for substances with known or suspected common modes of action, using whole effluent assessment methods to test chronic toxicity directly, coupled with specific site work to look for evidence of impacts). Such a study would need to be carefully designed to avoid the limitations imposed by testing methodology and current chemical monitoring and biological quality datasets as far as possible. The Environment Agency’s GC-MS monitoring suite could provide a basis for the identification of suitable sites. However, since the GC-MS is currently a semi-quantitative tool, significant improvement in analytical approaches would be necessary to avoid misleading conclusions and this is likely to be highly resource-intensive. Where suitable evidence exists, it is possible at a practical level to take account of additional toxicity caused by mixture effects by increasing the size of the assessment factor in the PNEC derivation for individual substances (eg as suggested by Backhaus et al. 2010). Further work could help provide examples, providing a flexible approach which considers the mode of action and highlight additional factors that may need to be considered on a case-by-case basis (eg comparing acute lethality due to narcosis versus chronic effects due to a specific mode of action such as endocrine disruption) (ECETOC 2007). However, given the limited empirical evidence for generic mixture effects on a wide scale, a blanket default approach to prospective risk assessment (eg under REACH) is expected to be inefficient and over-precautionary at the present time. In addition, the relevance of the other uncertainties that are addressed by the assessment factor would need to be considered, to ensure that any revised threshold was based on the best available scientific evidence (Chapman et al. 1998). In some cases, it may be possible to use toxic equivalency factors instead, provided that a suitable evidence base exists as is the case for dioxins and furans (Defra 2002) and other well-defined classes of environmental contaminants.


Finally, given the fundamental principles of mixture toxicology, it is important that environmental risk assessors maintain a close dialogue with specialists in mammalian toxicology and human risk assessment of chemicals. As an encouraging example, the flexible tiered approach to the cumulative risk assessment of chemical mixtures being developed by the WHO (2009) is built on clear scientific principles and pragmatism that may be useful not only for human risk assessment but also for environmental applications.

6 Acknowledgements

Our thanks to Leo Posthuma (RIVM The Netherlands) and Paul Whitehouse (EA Wallingford) for their help with the eco-epidemiology aspects of this report and also to Paola Cassanelli and Mike Roberts (Defra) for their valuable advice. Our thanks also to colleagues in the Department of Environment for Northern Ireland, the Environment Agency of England and Wales and the Scottish Environmental Protection Agency for supplying environmental quality data. For practical purposes, the report Annex is limited to examples of the large datasets provided.

7 References

ACHS (2010) Comment on the report of the Inter-departmental Group on Health Risks from Chemicals Chemical Mixtures: A framework for assessing risks to human health. UK Advisory Committee on Hazardous Substances. Available from: http://archive.defra.gov.uk/environment/quality/chemicals/achs/documents/achs-ighrc-chemicals-framework.pdf

Altenburger R, Backhaus T, Boedeker W, Faust M, Scholze M, Grimme LH (2000) Predictability of the toxicity of multiple chemical mixtures to Vibrio fischeri: mixtures composed of similarly acting chemicals. Environ Toxicol Chem 19:2341–2347

Backhaus T, Faust M, Scholze M, Gramatica P, Vighi M, Grimme LH (2004) Joint algal toxicity of phenylurea herbicides is equally predictable by concentration addition and independent action. Environ Toxicol Chem 23:258–264

Backhaus T, Blanck H, Faust M (2010) Hazard and Risk Assessment of Chemical Mixtures under REACH State of the Art, Gaps and Options for Improvement. Swedish Chemicals Agency report reference PM 3/10. 77pp. Available from: http://www.kemi.se/upload/Trycksaker/Pdf/PM/KemI_PM_3_10.pdf

Bjergager MBA, Hanson ML, Lissemore L, Henriquez N, Solomon KR, Cedergreen N (2011) Synergy in microcosms with environmentally realistic concentrations of prochloraz and esfenvalerate. Aquat Toxicol 101: 412-422

Bliss CI (1939) The toxicity of poisons applied jointly. Ann J App Biol 26: 585-615

Brian JV, Harris CA, Scholze M, Backhaus T, Booy P, Lamoree M, Pojana G, Jonkers N, Runnalls T, Bonfà A, Marcomini A, Sumpter JP (2005) Accurate prediction of the response of freshwater fish to a mixture of estrogenic chemicals. Environ Hlth Perspect 113:721–728

Brian JV, Harris CA, Runnalls TJ, Fantinati A, Pojana G, Marcomini A, Booy P, Lamoree M, Kortenkamp A, Sumpter JP (2008) Evidence of temperature-dependent effects on the estrogenic response of fish: implications with regard to climate change. Sci Total Environ 397: 72-81

Calamari D and Vighi M (1992) A proposal to define quality objectives for aquatic life for mixtures of chemical substances. Chemosphere 25:531-542


Cedergreen N, Christensen AM, Kamper A, Kudsk P, Mathiassen SK, Streibig J C, Sörensen H (2008) A review of independent action compared to concentration addition as reference models for mixtures of compounds with different molecular target sites. Environ Toxicol Chem 27:1621-1632

Chapman P M, Fairbrother A, Brown D (2009) A critical evaluation of safety (uncertainty) factors for ecological risk assessment. Environ Toxicol Chem 17: 99-108

Chevre N, Loepfe C, Singer H, Stamm C, Fenner K, Escher BI (2006) Including mixtures in the determination of water quality criteria for herbicides in surface water. Environ Sci Technol 40:426-435

Crane M, Kwok KWH, Wells C, Whitehouse P, Lui GCS (2007) Use of field data to support European Water Framework Directive Quality Standards for dissolved metals. Environ Sci Technol 41: 5014–5021

Defra (2002) Dioxins and dioxin-like PCBs in the UK environment. Consultation document. 98 pp

Defra (2010) Charting Progress 2: The State of the UK Seas. Defra report reference PB 13420, 166 pp (see also http://chartingprogress.defra.gov.uk/)

De Zwart D, Posthuma L (2005) Complex mixture toxicity for single and multiple species: Proposed methodologies. Environ Toxicol Chem 24: 2665–2676

EA (undated) GC-MS Target Based Screening (information note). National Laboratory Service. Environment Agency for England and Wales.

EA (2003) Derivation and Expression of Water Quality Standards – opportunities and constraints in adopting risk-based approaches in EQS setting (R&D Technical Report P2-157/TR). Environment Agency, Bristol. Available from: http://publications.environment-agency.gov.uk/pdf/SP2-157-TR-e-p.pdf

EA (2008a) Understanding ecological impacts in rivers in England and Wales and identifying their possible causes: Part 1 - The Effect and Probable Cause (EPC) method. Science Report SC030189/SR5. Environment Agency, Bristol. Available from: http://publications.environment-agency.gov.uk/pdf/SCHO1008BORM-e-e.pdf

EA (2008b) Understanding ecological impacts in rivers in England and Wales and identifying their possible causes: Part 2 - The GIS-based Weights of Evidence/Weighted Logistic Regression method. Science Report SC030189/SR6. Environment Agency, Bristol. Available from: http://publications.environment-agency.gov.uk/pdf/SCHO1008BORR-e-e.pdf

EA (2011a) Statistical approaches for distinguishing individual chemical toxicity thresholds in potentially complex mixtures. Poster available from: http://www.wca-environment.com/wp-content/uploads/2011/02/A0_Statistical-approaches-for-complex-mixtures.pdf

EA (2011b) Routine environmental monitoring in rivers - Supporting information for chemistry: Operational instruction 034_08. Issued 21/3/2011.

EC (2002) Working Document: Guidance Document on Aquatic Ecotoxicology in the context of the Directive 91/414/EEC. Sanco/3268/2001 rev.4 (final). European Commission. 17 October 2002

EC (2003) Technical Guidance Document on Risk Assessment in support of Commission Directive 93/67/EEC on Risk Assessment for new notified substances, Commission Regulation (EC) No 1488/94 on Risk Assessment for existing substances Directive, and 98/8/EC of the European Parliament and of the Council concerning the placing of biocidal products on the market. Institute for Health and Consumer Protection, European Chemicals Bureau. EUR 20418 EN/1. European Communities.

ECETOC (2007) Intelligent Testing Strategies in Ecotoxicology: Mode of Action Approach for Specifically Acting Chemicals. Technical report 102, European Centre for the Ecotoxicology and Toxicology of Chemicals, Brussels, Belgium, 145pp. Available from www.ecetoc.org


ECHA (2008a) Guidance on information requirements and chemical safety assessment. Part A: Introduction to the Guidance Document. European Chemicals Agency, Helsinki. 46 pp. Available from:

http://guidance.echa.europa.eu/docs/guidance_document/information_requirements_part_a_en.pdf

ECHA (2008b) Guidance on information requirements and chemical safety assessment. Chapter R.10: Characterisation of dose [concentration]-response for environment. European Chemicals Agency, Helsinki. 64 pp. Available from:

http://guidance.echa.europa.eu/docs/guidance_document/information_requirements_r10_en.pdf?vers=20_08_08

Faust M, Altenburger R, Backhaus T, Blanck H, Boedeker W, Gramatica P, Hamer V, Scholze M, Vighi M, Grimme LH (2001) Predicting the joint algal toxicity of multi-component s-triazine mixtures at low-effect concentrations of individual toxicants. Aquat Toxicol 56:13-32

Grothe DR, Dickson KL, Reed-Judkins DK (1996) Whole Effluent Toxicity Testing: An Evaluation of Methods and Prediction of Receiving Environment Impacts. Pensacola, FL, USA: SETAC.

Hering D, Borja A, Carstensen J, Carvalho L, Elliott M, Feld CK, Heiskanen AS, Johnson RK, Moe J, Pont D, Solheim AL, van de Bund W (2010) The European Water Framework Directive at the age of 10: A critical review of the achievements with recommendations for the future. Sci Total Environ 408: 4007-4019

IGHRC (2009) Chemical Mixtures: A Framework for Assessing Risk to Human Health (CR14). Institute of Environment and Health, Cranfield University, UK. ISBN 978-1-899110-44-5. Available from:

http://ieh.cranfield.ac.uk/ighrc/Chemical%20Mixture%20Final%20May%202009.pdf

Jobling S, Nolan M, Tyler CR, Brighty G, Sumpter JP (1998) Widespread sexual disruption in wild fish. Environ Sci Technol 32: 2498-2506

Junghans M, Backhaus T, Faust M, Scholze M, Grimme LH (2006) Application and validation of approaches for the predictive hazard assessment of realistic pesticide mixtures. Aquat Toxicol 76:93-110

Kajiwara Y, Nobuhisa K, Kiwao K (2009) Nationwide study of dioxins in the freshwater fish Carassius auratus (gibelio) and langsdorfii (crucian carp) in Japan: Concentrations and estimation of source contribution ratios. Chemosphere 69: 1177-1187

Kortenkamp A (2007) Ten years of mixing cocktails: a review of combination effects of endocrine-disrupting chemicals. Environ Health Perspect 115 Suppl 1:98-105

Kortenkamp A, Backhaus T, Faust M. (2009) State of the Art Report on Mixture Toxicity. Report for Directorate General for the Environment of the European Commission

Laetz CA, Baldwin DH, Collier TK, Hebert V, Stark JD, Scholz NL (2008) The synergistic toxicity of pesticide mixtures: implications for risk assessment and the conservation of endangered Pacific salmon. Environ Health Perspect 117(3):348-353

Lawler SP, Dritz DA, Johnson CS, Wolder M (2008) Does synergized pyrethrin applied over wetlands for mosquito control affect Daphnia magna zooplankton or Callibaetis californicus mayflies? Pest Management Science 64: 843-847

Liess M, Schäfer RB, Schriever CA (2008) The footprint of pesticide stress in communities—Species traits reveal community effects of toxicants. Sci Total Environ 406: 484-490

Matthiessen P, Whale GF, Rycroft RJ, Sheahan DA (1988) The joint toxicity of pesticide tank-mixes to rainbow trout. Aquat Toxicol 13: 61-75

OECD (1995) Guidance Document for Aquatic Effects Assessment. OECD Series on Testing and Assessment no 3. OECD, Paris 118 pp


OECD (2000) Guidance Document on Aquatic Toxicity Testing of Difficult Substances and Mixtures. OECD Series on Testing and Assessment no 23. OECD, Paris 53pp.

OSPAR Commission (2005) Whole Effluent Assessment. Report. OSPAR Publication Number: 2005/219. 123pp.

Posthuma L, de Zwart D (2006) Predicted effects of toxicant mixtures are confirmed by changes in fish species assemblages in Ohio, USA rivers. Environ Toxicol Chem 25: 1094–1105

Silva E, Rajapakse N, Kortenkamp A (2002) Something from “nothing”—eight weak estrogenic chemicals combined at concentrations below NOECs produce significant mixture effects. Environ Sci Technol 36:1751–1756

Sumpter JP, Johnson AC, Williams RJ, Kortenkamp A, Scholze, M (2006) Modelling effects of mixtures of endocrine disrupting chemicals at the river catchment scale. Environ Sci Technol 40:5478-5489

Syberg K, Jensen TS, Cedergreen N, Rank J (2009) On the use of mixture toxicity assessment in REACH and the Water Framework Directive: a review. Human Ecological Risk Assess 15: 1257-1272

Thorpe KL, Hutchinson TH, Hetheridge MJ, Scholze M, Sumpter JP, Tyler CR (2001) Assessing the biological potency of binary mixtures of environmental estrogens using vitellogenin induction in juvenile rainbow trout (Oncorhynchus mykiss). Environ Sci Technol 35:2476–2481

Thorpe KL, Cummings RI, Hutchinson TH, Scholze M, Brighty G, Sumpter JP, Tyler CR (2003) Relative potencies and combination effects of steroidal estrogens in fish. Environ Sci Technol 37:1142–1149

US EPA (2000) Guidance for Conducting Health Risk Assessment of Chemical Mixtures. Available from http://www.epa.gov/ncea/pdfs/mixtures.pdf

Vighi M, Altenburger R, Arrhenius Å, Backhaus T, Boedeker W, Blanck H, Consolaro F, Faust M, Finizio A, Froehner K, Gramatica P, Grimme LH, Grönvall F, Hamer V, Scholze M, Walter H (2003) Water quality objectives for mixtures of toxic chemicals: problems and perspectives. Ecotoxicol Environ Safety 54:139-150

Vighi M, Calamari D (1996) Quality objectives for aquatic life: the problem of mixtures of chemical substances. Human Ecological Risk Assess 2: 412-418

Walter H, Consolaro F, Gramatica P, Scholze M, Altenburger R (2002) Mixture toxicity of priority pollutants at No Observed Effect Concentrations (NOECs). Ecotoxicol 11:299-310

Wan Y, Jones PD, Holem RR, Khim JS, Chang H, Kay SP, Roark SA, Newsted JL, Patterson WP, Giesy JP (2010) Bioaccumulation of polychlorinated dibenzo-p-dioxins, dibenzofurans, and dioxin-like polychlorinated biphenyls in fishes from the Tittabawassee and Saginaw Rivers, Michigan, USA . Science of The Total Environment 408: 2394-2401

Whitehouse P, Johnson I, Forrow DM, Chubb C (2004) A regulatory framework for controlling effluent discharges using toxicity testing in the UK. Ecotoxicol 13: 399-411

WHO (2009) Risk assessment of combined exposures to multiple chemicals: a WHO/IPCS Framework. Published under the joint sponsorship of the World Health Organization, the International Labour Organization and the United Nations Environment Programme, and produced within the framework of the Inter-Organization Programme for the Sound Management of Chemicals. 17pp


Annex 1: Chemicals monitored in UK surface waters

Routine monitoring of chemicals in UK surface waters takes place for a variety of reasons, with both international and national drivers (e.g. Dangerous Substances Directive (76/464/EEC), Nitrates Directive (91/676/EEC) and Oslo & Paris (OSPAR) Commission). These determine the types of sites to be sampled, sampling frequencies and specific substances that need to be investigated. Full details for England and Wales are provided in EA (2011b). One of the primary reasons for collecting data is for the purposes of river basin planning under the EU Water Framework Directive. Monitoring is required to classify water bodies according to their ecological and chemical status, refine risk assessments to improve our understanding of possible threats to the environment, and guide improvement measures, as follows: • surveillance - to identify long term changes at a network of fixed sites; • operational - tailored monitoring to help classify water bodies that are at risk of failing to

meet objectives; • investigative - to assess why a water body is failing to achieve its objectives and decide

what action is needed; • groundwater monitoring - to determine the quantitative status of groundwater bodies; and • protected area monitoring - in surface and ground waters used for the abstraction of

drinking water; habitat and species protection areas designated under the Habitats Directive, or any other protected area established by EU legislation.

A range of parameters are monitored for each water body, including:

- biology (phytoplankton, diatoms, macrophytes, invertebrates and fish); - hydromorphology; - physico-chemical; and - 22 priority substances, 11 priority hazardous substances (as defined in the Water

Framework Directive) (see Table A.1) and around 30 specific pollutants including metals (arsenic, chromium (III) and (VI), copper, iron and zinc), other inorganics (chlorine, cyanide), pesticides (e.g. aldrin, certain pyrethroids, DDT, diazinon) and other organic substances (e.g. certain chlorinated hydrocarbons, phenol and toluene).

Some additional substances are monitored under other schemes, but there is a lot of commonality with the Water Framework Directive. Despite the focus on specific known substances of high concern, routine monitoring arrangements can only provide a partial picture of the extent of chemical contamination at any particular site. In addition, some of the substances that are included in monitoring suites are no longer used in any significant amounts, so concentrations are frequently reported as not detected. The Environment Agency has recently developed a gas chromatography-mass spectroscopy (GC-MS) method that is able to detect a much larger number of substances than those monitored to comply with the Water Framework Directive (EA, undated). It is used routinely in England and Wales for about 3,000 groundwater samples a year (and in future marine water samples will also be screened). It is also used occasionally for effluents and incident investigations, and it can also be applied to sediments.


Table A.1 Water Framework Directive Priority and Priority Hazardous Substances

Name CAS number Name CAS number

Alachlor 15972-60-8 Lead and its compounds 7439-92-1 Anthracene 120-12-7 Mercury and its compounds 7439-97-6 Atrazine 1912-24-9 Naphthalene 91-20-3 Benzene 71-43-2 Nickel and its compounds 7440-02-0 Pentabromodiphenyl ether (BDE congener numbers 28, 47, 99, 100, 153 and 154)

32534-81-9 Nonylphenols 25154-52-3 104-40-5

Cadmium and its compounds 7440-43-9 Octylphenols 1806-26-4 140-66-9

Chloroalkanes, C10-13 85535-84-8 Pentachlorobenzene 608-93-5 Chlorfenvinphos 470-90-6 Pentachlorophenol 87-86-5 Chlorpyrifos (Chlorpyrifos-ethyl)

2921-88-2 Polyaromatic hydrocarbons (Benzo(a)pyrene) (Benzo(b)fluoranthene) (Benzo(g,h,i)perylene) (Benzo(k)fluoranthene) (Indeno(1,2,3-cd)pyrene)

not applicable 50-32-8 205-99-2 191-24-2 207-08-9 193-39-5

1,2-Dichloroethane 107-06-2 Simazine 122-34-9 Dichloromethane 75-09-2 Tributyltin compounds

Tributyltin-cation not applicable 36643-28-4

Di(2-ethylhexyl)phthalate (DEHP)

117-81-7 Trichlorobenzenes 12002-48-1

Diuron 330-54-1 Trichloromethane (chloroform) 67-66-3 Endosulfan 115-29-7 Trifluralin 1582-09-8 Fluoranthene 206-44-0 Hexachlorobenzene 118-74-1 Hexachlorobutadiene 87-68-3 Hexachlorocyclohexane 608-73-1 Isoproturon 34123-59-6

The method is capable of giving semi-quantitative concentration data for over 860 discrete substances present at low levels in a single sample: the detection limit is 0.01 µg/l for three quarters of the compounds. This list is growing year on year, and a more tailored adaptation can identify up to 190,000 chemicals. The obvious attraction of this approach is that it can include large numbers of chemicals that are not routinely detected by statutory schemes but for which there is scientific evidence that they may be affecting the ecological quality of a water body. The main drawback of the GC-MS method is that it does not give accurate concentration data: concentrations are generally within a factor of ten of the reported measurement (although they are often much closer). As an example, Table A.2 presents data for substances that were detected in a real sample of trade effluent. This effluent contained at least 78 discrete organic substances, at a total concentration of around 215 µg/l (0.2 mg/l). However, just four substances accounted for almost two thirds of the measured chemical contaminant load (the PVC plasticiser DEHP (22 µg/l), and three pesticides: chlorpropham (80 µg/l), tebuconazole (14 µg/l) and chlorotoluron (12 µg/l)). Other substances may have been present but not looked for (because they are not in the target database), or could be below the limit of detection of the method. It should also be noted that the final concentrations in the receiving water will depend on the efficiency of the wastewater treatment method (which will vary for different substances), partitioning processes (e.g. with suspended matter) and the dilution available.


Table A.2 Substances detected in an anonymised trade effluent using GC-MS

CAS no. Substance Approx. conc. (µg/l)

CAS no. Substance Approx. conc. (µg/l)

13-11-3 Dimethyl phthalate 0.5 206-44-0 Fluoranthene 5.0 50-32-8 Benzo[a]pyrene 1.5 207-08-9 Benz[k]-

fluoranthene 1.6

51-28-5 2,4-Dinitro-phenol 0.4 208-96-8 Acenaphthylene 0.1 53-70-3 Dibenz[a,h]-anthrancene 0.5 218-01-9 Chrysene 2.0 56-55-3 Benz[a]-anthracene 2.0 330-55-2 Linuron 1.7 58-08-2 Caffeine 1.0 534-52-1 2-Methyl-4,6-

dinitrophenol 0.2

75-25-2 Bromoform 0.1 886-50-0 Terbutryne 0.3 75-27-4 Bromodichloro-methane 0.1 1194-65-6 2,6-Dichloro-

benzonitrile 0.1

81-84-5 Naphthalic anhydride 0.4 1689-84-5 Bromoxynil 0.3 83-32-9 Acenaphthene 0.1 1698-60-8 Pyrazon 6.0 84-65-1 9,10-Anthraquinone 0.1 1702-17-6 Clopyralid 3.8 84-74-2 Di-n-butyl phthalate 4.0 1861-32-1 Chlorthal-

dimethyl 0.2

85-01-8 Phenanthrene 1.0 1912-24-9 Atrazine 0.4 86-73-7 Fluorene 0.1 1918-16-7 Propachlor 0.4 88-74-4 2-Nitroaniline 0.2 2303-16-4 Diallate 0.3 91-20-3 Naphthalene 0.1 2303-17-5 Tri-allate 3.8 91-64-5 Coumarin 0.1 3481-20-7 2,3,5,6-Tetra-

chloroaniline 0.1

91-57-6 2-Methyl-naphthalene 0.02 5989-27-5 d-Limonene 0.1 95-48-7 o-Cresol (2-Methylphenol) 0.3 13194-48-4 Ethoprophos 0.2 95-53-4 o-Toluidine 0.3 15545-48-9 Chlorotoluron 12 98-86-2 Acetophenone 0.2 16118-49-3 Carbetamide 0.3 100-02-7 4-Nitrophenol 0.2 23950-58-5 Propyzamide 1.4 100-42-5 Styrene 0.1 23103-98-2 Pirimicarb 0.2 100-51-6 Benzyl alcohol 0.04 21087-64-9 Metribuzin 0.3 101-21-3 Chlorpropham 80 26225-79-6 Ethofumesate 0.5 105-67-9 2,4-Dimethyl-phenol 0.1 34123-59-6 Isoproturon 5.0 106-47-8 4-Chloroaniline 4.7 40487-42-1 Pendimethalin 0.6 108-42-9 3-Chloroaniline 5.0 41394-05-2 Metamitron 1.0 108-68-9 3,5-Dimethyl-phenol 0.1 57837-19-1 Metalaxyl 0.2 108-90-7 Chlorobenzene 0.1 52888-80-9 Prosulfocarb 0.3 110-86-1 Pyridine 0.4 67564-91-4 Fepropimorph 0.1 117-81-7 Bis(2-ethyl-hexyl)

phthalate (DEHP) 22 83164-33-4 Diflufenican 1.0

120-12-7 Anthracene 1.3 85509-19-9 Flusilazole 0.4 124-48-1 Chlorodibromo-methane 0.1 107534-96-3 Tebuconazole 14 129-00-0 Pyrene 3.0 110488-70-5 Dimethomorph 1.2 132-64-9 Dibenzofuran 0.1 113096-99-4 Cyproconazole 0.8 191-24-2 Benzo[ghi]-perylene 1.5 121552-61-2 Cyprodinil 0.1 193-39-5 Indeno[1,2,3-cd]pyrene 3.8 142459-58-3 Flufenacet 7.0 205-99-2 Benz[b]-fluoranthene 2.8 188425-85-6 Boscalid 4.0

The original purpose for commissioning this report was to review chemical monitoring data sets collected by UK national regulatory authorities to see if any important or interesting correlations could be found at sites with differing ecological status, in the context of the mixtures of substances that are present. After discussion and agreement between Defra and the report authors, it was agreed that the UK and US eco-epidemiology case studies summarised in the main text (Section 4.5.1) made such an analysis superfluous in many respects. The large number of ‘non-detects’ reported for specific chemical determinands would also have a confounding effect on detailed statistical correlations which could not have been properly addressed within the scope and timeframe of providing this report to Defra. Indeed, the effluent example given in Table A.2 illustrates the complexity of working out which substances might possibly be involved in


ecotoxicological interactions in the wider environment. It is only a snapshot for one particular location, and the concentrations may vary substantially over time, with some substances featuring on some occasions but not others. The wide variety of different chemical types represented by the substances also shows how difficult it is even to screen for possible common modes of action (e.g. using the approaches proposed by ECETOC (2007)). In general terms, it might be expected that those chemicals that are designed to be biologically active (i.e. pesticides and pharmaceuticals) could be the most relevant to consider in this way (Thorpe et al. 2003; Laetz et al. 2008), but where other substances with non-specific modes of action form the greater part of a sample, they might also have important collective effects. In addition, some general chemicals can have unexpected properties. The results of correlation exercises where only a small number of substances have been considered, without supporting toxicological investigations, should therefore be treated with a large degree of caution. Any apparent observed toxicity could be caused by other substances (or non-chemical factors) that have not been considered, or by monitored chemicals that occur in a given sample at concentrations below the analytical detection limit.


About us Cefas is a multi-disciplinary scientific research and consultancy centre providing a comprehensive range of services in fisheries management, environmental monitoring and assessment, and aquaculture to a large number of clients worldwide. We have more than 500 staff based in two laboratories, our own ocean-going research vessel, and over 100 years of fisheries experience. We have a long and successful track record in delivering high-quality services to clients in a confidential and impartial manner. Cefas Technology Limited (CTL) is a wholly owned subsidiary of Cefas specialising in the application of Cefas technology to specific customer needs in a cost effective and focussed manner. CTL systems and services are developed by teams that are experienced in fisheries, environmental management and aquaculture, and in working closely with clients to ensure that their needs are fully met. For more information see http://www.cefas.defra.gov.uk.

Customer focus With our unique facilities and our breadth of expertise in environmental and fisheries management, we can rapidly put together a multi-disciplinary team of experienced specialists, fully supported by our comprehensive in-house resources. Our existing customers are drawn from a broad spectrum with wide ranging interests. Clients include: • international and UK government departments • the European Commission • the World Bank • Food & Agriculture Organisation of the United Nations • oil, water, chemical, pharmaceutical, agro-chemical, aggregate and marine industries • non-governmental and environmental organisations • regulators and enforcement agencies • local authorities and other public bodies We also work successfully in partnership with other organisations, operate in international consortia and have several joint ventures commercialising our intellectual property.

HEAD OFFICE Centre for Environment, Fisheries & Aquaculture Science Pakefield Road, Lowestoft, Suffolk NR33 0HT UK Tel +44 (0) 1502 562244 Fax +44 (0) 1502 513865 For more information see http://www.cefas.defra.gov.uk

Centre for Environment, Fisheries & Aquaculture Science Barrack Road The Nothe Weymouth Dorset DT4 8UB UK Tel +44 (0) 1305 206600 Fax +44 (0) 1305 206601

Date post:	15-Jun-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

Environmental Risk Assessment of Chemical Mixtures in...

Documents