Department for Environment, Food and Rural Affairs Research project final report

Project title Review of the literature on the concentrations of contaminants in that have the potential to lead to significant concentrations of contaminants in food.

Sub-project E of Defra Project SP1605: Studies to support future Policy

Defra project code SP1605

Contractor SKM Enviros organisations Rothamsted Research

ADAS

Report authors Steve McGrath ([email protected]), Fiona Nicholson, Fangjie Zhao

Project start date November 2009 Project end date April 2010

Glossary

ADI Acceptable daily intake Ag Silver As Arsenic Au Gold BaP Benzo-a-pyrene – a type of PAH BFR Brominated flame retardants BW Body weight Cd Cadmium Ce Cerium CEC Commission of the European Communities CF Concentration factor – a measure of the concentration in plant material relative to the concentration in soil CLEA Exposure Assessment Contamination Presence of a substance above concentrations usually regarded as ambient (for xenobiotic organic compounds this is regarded as any detectable concentration) COT Committee on Cr Chromium Cu Copper CP Chlorinated paraffin Defra Department for Environment, Food and Rural Affairs DoE Department of the Environment DW Dry weight EFSA European Food Standards Agency F Fluorine FAO Food and Agriculture Organisation FW Fresh weight HCB Hexachlorobenzene Hg JECFA Joint FAO/WHO Expert Committee on Food Additives JMPR Joint FAO/WHO Meeting on Residues Koc Organic carbon-water partition coefficient – a measure of adsorption to organic matter KOW Octanol-water partition coefficient - a measure of liphophilicity ML Maximum allowed concentrations Mn Manganese Mo Molybdenum MPC Maximum permitted concentration MRL Maximum residue limit NP Nonylphenol Ni Nickel PAH Polycyclic aromatic hydrocarbon Pb Lead PCB Polychlorinated biphenyl PCDD/F Polychlorinated dibenzo-p-dioxin / polychlorinated dibenzofuran PCP Pentachlorophenol PFA Perfluoroalkyl substance pg (picogram) One trillionth (10-12) of a gram PFOS Perfluorooctane sulfonate PFOA Perfluorooctanoic acid PTWI Provisional tolerable weekly intake

Page 1

RA Risk Assessment Sb Antimony Se Selenium SGV Soil guideline values Si Silicon Sn Tin SOM Soil organic matter TDI Tolerable daily intake Ti Titanium Tl Thallium TWI Tolerable weekly intake US EPA US Environmental Protection Agency V Vanadium WHO World Health Organisation WHO-TEQ WHO Toxic Equivalent Quotient. – weighted toxicity factors for PCDD/Fs relative to the most toxic compound, set by the WHO. Zn Zinc Zr Zirconium Σ Sum of the compounds in a particular chemical group

Page 2

1. Introduction

This review is of the literature related to the concentrations of contaminants in soils that have the potential to lead to significant concentrations in food. Contamination is taken to denote the presence of, in this case, metals or organic contaminants above background levels. But this does not necessarily mean that they are having a deleterious effect. The title of this review implies that at some concentrations contaminants can be associated with detrimental effects on organisms (man in this case). It is difficult to establish casual relationships between exposure to contaminated soil and actual health impacts (Defra, 2009a).These are potential effects, as it is obviously not desirable to allow concentrations in food to become high enough to have actual effects. Therefore, various agencies have set limits in food and feed that are less than the levels at which effects on animal and human health can occur. These are predicted upper allowable limits and can change with time according to new knowledge and following review by the responsible agencies. We take concentrations that exceed these limits in food or feed to be “significant” and those below are insignificant. However, it should be noted that there are other sources of environmental contamination of the food chain and that soil is only one potential source.

A site is not be legally defined as ‘contaminated land’ under Part 2A of the Environment Protection Act 1990, unless the levels of contaminants present (and the routes by which humans or the environment are impacted) are sufficient for there to be a significant possibility of significant harm. We have used information that is relevant to situations in which the soil is likely to be contaminated and there is a possibility of food being grown (e.g. urban gardens and allotments). However, we exclude known ‘contaminated land’, including industrial sites, sites or any other sites that are locally impacted by specific polluting events (i.e. the focus of this review is on diffuse ). Sites which are in use for producing food and agricultural land that has had sludge or wastes applied are included.

As noted in the Soil Strategy for England, continued atmospheric deposition of contaminants and the increase in the spreading of organic and inorganic materials which may contain contaminants to land is an area of growing concern (Defra, 2009b). Materials used on land include organic wastes (e.g. composts, animal manure, sewage sludge) and inorganic by- products (e.g. recycled gypsum). These materials are spread on land because they can increase the levels of organic matter in soil, reduce fertiliser or lime requirements, and provide the associated benefit of reduced landfilling or . The nutrients and organic matter can have important agricultural and ecological benefits. However, these materials can also contain significant concentrations of contaminants, such as metals or persistent organic contaminants, which can accumulate in soil following repeated applications (Nicholson et al, 2008). The input of metals from various sources has been studied in the UK (Nicholson et al, 2003, 2008), as has the accumulation of metals in soils receiving sewage sludge, animal manures and composts (McGrath 1987, Gibbs et al, 2006, Chaudri et al, 2007). However, the inputs and persistence of organic contaminants is not so well characterized. New and future threats may emerge, for example, nanoparticles are noted in this respect in the Soil Strategy (Defra, 2009b).

The presence of contaminants at high concentrations can adversely affect soil organisms which are necessary for healthy soil functions, as well as plants, animals and humans. The contaminants covered in this review are detailed in Sections 2 and 3. Both the chemical properties and the biological behavior of contaminants are important to their likely transfer to food. The following sections give an assessment of how the contaminants behave in soils under different conditions, how they transfer to animals and plants used as food, and under

Page 3 what conditions they are likely to breach the maximum allowed concentrations (ML) in food consumed by humans. Many of the strongest emission sources of contaminants in soils have been controlled in the last 25 years through environmental regulation, but because contaminants can persist for long periods in soil, and their continued arrival in soils via pathways such as atmospheric deposition and the spreading of organic and inorganic materials on land, soils can continue to contain elevated concentrations of contaminants or even increasing concentrations.

2. Metals/metalloids

2.1 Introduction

Metals and metalloids often accumulate in surface soil, and can be taken up by crops to different extents. Some of these are essential to plants and/or animals (e.g. Zn, Cu, Mn, Ni, Cr, Co, Mo) and deficiencies can occur when either the concentrations in soil are small or when the soil conditions limit their bioavailability. These can be toxic when present in large amounts, but this is unlikely to occur through the dietary route via soil and agriculture for reasons discussed below. Other elements (such as Hg, Cd, Pb, Tl and As) are not thought to be required biologically and are often toxic when they are present in sufficient concentrations.

2.2 Limits in food or the diet

This review concentrates on contaminants that build up in soil and have the potential to lead to significant accumulation in food. To do this it is necessary to examine the links between contaminant concentrations in soil and the presence of significant concentrations or intakes in food. Food in this sense should include the transfer from soil to plants to animals and animal products which are consumed, as well as plant foods that are consumed directly by humans. Direct transfer of contaminants from soil to animal (via soil ingestion) and then to humans is a potential route for contaminant transfer which is given some consideration below. In addition to concentrations, the amounts of specific foodstuffs consumed are important (for example As in rice; Meharg et al, 2009). We reviewed and compiled the existing maximum limit concentrations in food (ML; CEC 2001). Where they exist, these can be used as indicators of “significant” amounts of contaminants in food.

Among the toxic metals/metalloids, food (or animal feeding stuff) limits are in place for Cd, F, Pb, Hg and Sn (Annexes 1, 2, 3) or being considered for As. Of these, F is only present in animal feed limits because animals were specifically exposed in the past near to strong point sources such as brickworks and aluminium smelters. Similarly, Sn is regulated only in canned foods because there is a specific, acute risk of human exposure to tin from damaged or unlacquered cans. Exposure by this route has been minimised by introducing controls. Mercury is most harmful as methyl mercury, which is generally not found in terrestrial foods. It is the commonest form in fish, which is why there are limits in seafood only. For these reasons, F, Hg and Sn will not be discussed in any detail below.

It should be noted that the existing limits for Cd and Pb are under review, and introduction of limits for As in rice is being considered. The main source of Cd for non-smokers is food. For Cd, the EFSA Panel on Contaminants in the Food Chain (EFSA, 2009a) recommended that the PTWI for this element should decrease by a factor of almost 3. The European Commission is considering decreasing the ML in various food stuffs, for example, a proposal to amend Food Regulation (EC) No. 1881/2006 suggests that the maximum concentration in wheat may drop by half (FSA, 2009a). This would affect all of the evaluations of soil to food transfers, for example in section 2.6 below. For As, the diet is the main source of exposure in the EU and a recently published opinion was that this should be decreased (EFSAb, 2009)

Page 4 as much as possible. In order to do that, more data are required on the chemical species of As in different foodstuffs and on the relative toxicity of each compound. Lead is also regarded as an element that intake should be as low as reasonably practicable because EFSA could not establish a threshold value for developmental neurotoxicity (FSA, 2009b).

2.3 Behaviour in soil

Soil conditions can affect the build up and losses of contaminants and, therefore, their bioavailability and likely transfer through the food chain (Chaudri et al, 2007, Rooney et al, 2006, Li et al, 2008). Understanding the soil conditions which affect plant uptake is important in two respects: a) these can be used to define soil-specific threshold values rather than a “one fits all” approach (Chaudri et al, 2007, Rooney et al, 2006) and, b) soil conditions that can be manipulated (e.g. pH) can form the basis of controls on the risk of contaminants entering the food chain in significant quantities. Metals which exist mainly as cationic (positively charged) species tend to become more available at acid pHs and precipitate and are bound by soils at under high pH conditions. Those metals or metalloids which form anionic (negatively charged) species usually react in the opposite way, being bound by positively bound sites e.g. on oxides at low pH and are less bound and, therefore, more available at high pHs. Other conditions can affect elements to variable extents, for example, high organic matter contents can bind metals like Cu and Pb strongly. Finally, the availability of oxygen (usually mediated by the water status of soil) affects the chemical form and binding of redox sensitive metals such as Mn, As and Cr.

2.4 Concentrations in soils

Extensive data exist on the concentrations of some metals and metalloids in some 5,600 soils of England and Wales in the National Soil Inventory (NSI) (McGrath and Loveland 1992). However, some potential contaminants are missing from these data (e.g. As, Se, Sb and Tl). The UK soils and herbage survey (Ross et al., 2007) gives analyses for eleven metals and metalloids in soil and herbage from 122 rural and 29 urban sites in the UK, but no quantitative data for Se, Sb and Tl. It would be useful to know from the point of view of toxicity and deficiency of elements whether the concentrations of metals and metalloids in soil are changing with time, for example due to , deposition from the atmosphere or to disposal of materials on agricultural land. These data do not exist on a national scale over a significant time interval (Ross et al., 2007).

Due to their potential toxicity to plants, animals and humans at high concentrations, soil maximum limit values for some metals have been published for the UK for the application of sewage sludge to agricultural soils. These are derived from the EU Directive on the use of sewage sludge on agricultural land (Tables 1 and 2). However, there is a lack of clarity on how values are derived in relation to human (and animal) health. Although the values apply only to land receiving sewage sludge, they tend to be used as benchmarks for example for land receiving bulky materials such as composts. The fact that a limit to the build-up of metals in topsoil exists constitutes another form of control which helps to prevent transfer of amounts of metals that could otherwise cause danger to humans. Similar soil maximum limit values have been published for grassland receiving sewage sludge; again, this will help limit the degree of transfer to animals (Table 2).

Page 5

Table 1. Maximum permissible concentrations of potentially toxic elements in the 0-25 cm depth of soil after application of sewage sludge, and maximum annual rates of addition (DoE, 1996).

PTE Maximum permissible in soil (mg/kg) Maximum permissible average annual rate of PTE a pH pH pH pH addition over a 10 year period (kg/ha) 5.0<5.5 5.5<6.0 6.0‐7.0 >7.0 Zinc 200 200 200 300 15 Copper 80 100 135 200 7.5 Nickel 50 60 75 110 3 For pH 5.0 and above Cadmium 2 0.15 Lead 300 15 Mercury 1 0.1 bChromium 400 15 bMolybdenum 4 0.2 bSelenium 3 0.15 bArsenic 50 0.7 Fluoride 500 20 aThe increased permissible PTE concentrations in soils of pH greater than 7.0 apply only to soils containing more than 5% calcium carbonate. b These are not subject to the provisions of Directive 86/278/EEC.

Page 6

Table 2. Maximum permissible concentrations of potentially toxic elements in soil under grass after application of sewage sludge when samples are taken to a depth of 7.5 cm (DoE, 1996).

PTE Maximum permissible concentration in soil (mg/kg) pH pH pH pHa Zinc 200 200 200 300 Copper 130 170 225 330 Nickel 80 100 125 180 For pH 5.0 and above Cadmium 3 Lead 300 Mercury 1.5 bChromium 600 bMolybdenum 4 bSelenium 5 bArsenic 50 *Fluoride 500 a The increased permissible PTE concentrations in soils of pH greater than 7.0 apply only to soils containing more than 5% calcium carbonate. b These are not subject to the provisions of Directive 86/27/EEC.

In addition to the above, Soil Guideline Values (SGVs) for some metals/metalloids have been published for England and Wales, which are cautious screening values that can be used to indicate when land would definitely not be classified as contaminated land (and as such can be used in the preliminary assessment of land suspected to be contaminated) (Environment Agency 2009a). The Environment Agency have published a framework approach that describes how the risks to human health from land contamination can be assessed, using SGVs for individual, or groups of similar, chemicals. SGVs are guidelines on the level of long-term human exposure to individual chemicals in soil that are tolerable or pose a minimal risk to human health. Although this report does not consider contaminated land, these are presented here because they are based on modelling and specifically aim to avoid potential effects on human health. The SGV values in Table 3 are based on a sandy soil described by the Environment Agency (2009a) and will vary according to soil type. The original source documents give full guidance on how these values should be applied, and here the values relevant to this review are those applied to residential or allotment land uses.

Page 7

Table 3. Soil Guideline values for preliminary assessment of contaminated land (Environment Agency 2009a)1.

All values mg/kg dry weight soila Element Residen Allotment Arsenic (inorganic As) 32 l 43 Cadmium 10 1.8 Mercury (inorganic) 170 80 Nickel 130 230 Selenium 350 120

1 Based on a sandy loam soil with 6% SOM as defined in Environment Agency (2009b).

2.5 Plant uptake and concentrations in food crops

A soil-plant barrier concept has been developed for metals and metalliods (Basta et al, 2005). This assumes that the mobility of these elements is influenced by soil or plant “barriers” that can limit transmission of many metals through the food chain either due to soil chemical processes that limit solubility (soil barrier) or by plant senescence due to phytotoxicity (plant barrier). Applying this concept, the soil barrier is thought to limit transmission of trace elements, including Ti, Si, Zr, Au, Ag, V, and Hg, that are very insoluble and/or strongly adsorbed in soil or in roots (Basta et al, 2005). Based on published soil to plant transfer factors (concentration in plant/concentration in soil) for a wide range of plant species, we would add Pb to this group with a soil barrier, although aerial deposition of soil or dust onto crops can result in transfer into the food chain, though not strictly taken up into the plant tissues. These elements are less likely to be bioaccumulated in plant shoots or edible storage organs to levels that constitute risk to humans than metals which are not affected by the soil barrier. On the other hand, Basta et al (2005) state that the plant barrier limits transmission of trace elements, including Zn, Cu, Ni, Mn, Cr, As and F because plant phytotoxicity occurs before they reach concentrations in the edible parts of plants considered harmful to consumers. However, in most soil conditions uptake of Cr and As are limited due to lack of bioavailability in soil (but note that this is only the case for As under aerobic soil conditions). The elements Cd, Mo, Se and Tl are thought to a certain extent to pass the soil– plant barrier that limits transmission of many trace elements through the soil–crop–animal food chain. In addition, paddy rice grown in anaerobic soil is efficient in accumulating As (Meharg et al, 2009). Section 2.6 will deal in more detail with Cd, the most important example of an element without a strong soil-plant barrier.

Se and Mo can also be relatively easily taken up by food crops. However, it is highly unlikely that their accumulation in food crops would cause toxicity to humans, because the concentrations of these are small in UK soils, with the exception of Mo accumulation in pasture herbage which can lead to Cu-deficiency in ruminants (Thomson et al., 1972). Approximately 0.5-1 billion people have inadequate Se intake, including the population in the UK (Combs, 2001; Rayman, 2002). A different example is Zn deficiency which approximately 2 billion people suffer from worldwide (White and Broadley, 2009). For Se, Zn and perhaps other essential trace elements the emphasis is to increase their concentrations and bioavailability in food crops to benefit human nutrition. However, for Se there is only about a 10-fold difference between the nutritionally necessary level and the harmful level.

Page 8

Thallium (Tl) is highly toxic to animals and humans and is readily taken up by some plants. A study in France showed very low Tl concentrations (< 4 μg/kg dry weight) in wheat and maize grain from crops grown in soils varying widely in Tl concentration, but high concentrations of Tl in winter oilseed rape (Brassica napus L) shoots and seeds with the soil to plant transfer factor > 1 (Tremel et al., 1997). Two other vegetable species, cabbage (Brassica oleracea cv capitata L.) and leek (Allium porrum L.) from a different location in France were also measured and showed transfer factors <1 (Tremel et al., 1997). The 2006 UK Total Diet Study showed low concentrations of Tl in foods and that, for average consumers, current dietary exposures to Tl are unlikely to be of toxicological concern (Food Standards Agency, 2009b). However, where soils containing elevated levels of Tl are used for growing food crops, based on the French study, there is potential for this to lead to increased uptake and accumulation. Further research should focus on vegetables (Brassica species) that are grown on soils with naturally elevated levels of Tl.

Table 4 summarises data from surveys of Cd, Pb and As concentrations in UK food crops. Chaudri et al. (1995) and Zhao et al. (2004) reported the ranges of Cd and Pb concentrations in wheat and barley grains collected representatively from major cereal- growing areas in the UK, which showed that >99% of the grain samples were below the current limits. If the Cd limit in wheat were to be reduced from 0.2 to 0.1 mg/kg fresh weight (0.113 mg/kg dry weight based on 88.8% dry matter in cereals) to reduce dietary exposure of Cd (Food Standards Agency, 2009a), 2.4% of the wheat grain would not be compliant. In a paired soil-crop survey which specifically included some land with elevated levels of Cd, Adams et al. (2004) found 4.5% of the wheat grain samples exceeding the current Cd limit. A survey of 40 barley grain samples collected in 2000 from Scotland showed anomalously high concentrations of Pb, with a quarter of the samples exceeding the limit (Zhao et al., 2004). However, it appears that these samples were surface contaminated during either sample collection or storage, as a brief washing of the grain decreased the Pb concentration markedly. Overall, most of the wheat and barley grain produced in the UK are compliant with the current Cd and Pb limits, although exceedance can occur if crops are grown at sites with elevated concentrations, or surface contamination of Pb can occur during storage and processing. Grain exceeding the limit would be diluted with other loads with low concentrations, and it is unlikely in this country that grain from a single source would be consumed exclusively.

There are no nationwide surveys of Cd and Pb concentrations in vegetables and potatoes produced from commercial farms in the UK. There are, however, surveys of vegetables grown in urban and a few rural allotments. Moir and Thornton (1989) reported wide ranges of Cd and Pb in vegetables grown in 94 urban allotments in England, with the maximal values recorded exceeding the current EU food limits for both metals (Table 4). The % exceedance was not reported. Leafy vegetables such as spinach and lettuce showed the highest accumulation of Cd. The generally high concentrations of Pb and Cd in this survey reflect the fact that urban soils tend to more contaminated with metals. A recent study of vegetables produced in 12 allotments in the UK (6 rural, 6 urban) reported generally low concentrations of Cd and Pb (Table 4) (Weeks et al., 2007).

The UK Soil and Herbage Survey conducted by the Environment Agency in 2002 (Ross et al., 2007) reported the ranges of As, Cd, Pb and other trace elements in herbage from 366 rural locations (Table 4). The maximal values are larger than the EU limits of these elements in animal feeding stuff, but the % exceedance is likely to be small judged from the means and medians.

There are no nationwide surveys of As concentration in UK cereal grains. Williams et al. (2007) reported As concentrations in wheat and barley grains collected from two regions in the UK (Table 4). While those from Scotland (mainly Fife) can be considered to be in the

Page 9 normal range, the concentrations in many of the samples collected from Southwest England (Devon and Cornwall) were high because soils had high (maximum 546 mg/kg) levels of As. The highest concentration in wheat grain reported by Williams et al. (2007) was 0.5 mg/kg, and this level could be of concern (if consumption levels are high and based on a single source of wheat) because As in wheat grain is present only in the inorganic forms (Zhao et al., 2010a). Where soils do not contain elevated levels of As (<20 mg/kg), the As concentrations found in wheat and barley grain are usually <0.05 mg/kg (Williams et al., 2007; Zhao et al., 2010b).

Page 10

Table 4. Concentrations of Cd, Pb and As in food crops (edible parts) in the UK.

Country Crop Element Concentration (mg/kg) % Note Reference DW or FW Minimum Maximum Mean Median n exceeding limit a UK Wheat Cd DW 0.004 0.31 0.043 0.037 1007 0.2 1982, 1992‐93, Nationwide, (Chaudri et al., 1995) representative UK Wheat Cd DW 0.01 0.62 0.077 0.056 162 4.5 1998‐2000, Nationwide, including some (Adams et al., 2004) sites with elevated Cd UK Barley Cd DW

UK Barley Pb DW

Page 11

2.6 Soil ingestion by livestock

Some metals may have the potential to transfer directly from the soil to be absorbed by grazing livestock (without passing through a plant), then transfer into the human food chain. Grazing livestock can ingest soil on its own or soil adhered to vegetation.

The amount of metal absorbed by this route depends on the quantity of soil ingested by the grazing animal, which varies depending on such factors as the livestock species, climate, grazing habits, time of year, sward cover and other factors such as stocking rates. It has been estimated that an average of around 6% of the livestock diet is soil (Fries, 1996; Stark and Hall 1992), although ingestion rates may be much higher under conditions of drought or over grazing, with Thornton and Abrahams, (1983) suggesting that up to 18% of the cattle diet may be soil. Concentrations of metals can be elevated when sewage sludge is applied to land due to adherence of the sludge to foliage and because the build up of metal is greatest in surface soil which animals are likely to consume. There is evidence that Cd and Pb accumulate in the liver and kidney, but not muscle tissues of animals that feed on land which has received a number of applications of sewage sludge (Hillman et al., 2003). The risk to humans from this route is reduced but not eliminated, for two main reasons: 1) the intake of offal is low in the average diet compared to muscle meat, and 2) to limit the risk of pathogen transfer to animals and amount of soil ingested in the case of sludged land, there is a no grazing or herbage harvesting period of 3 weeks after sludge application (Code of Practice for Agricultural Use of Sewage Sludge; DoE, 1996). However, worst cases scenarios could be more of an issue. An example would be people with a high consumption of offal, but this would be mitigated by the fact that it is unlikely that all of this particular dietary component would be derived from sludged land. Another case would be areas where the Code of Practice for Agricultural Use of Sewage Sludge was not adhered to, due to extreme weather conditions leading to poaching of soil, lack of herbage or poor management. Finally, the Code of Practice for Agricultural Use of Sewage Sludge does not apply to other routes of contamination of soils.

2.7 Models

A number of studies have attempted to establish empirical models to predict the transfer of metals from soil to crops, Cd being the most successful case as its transfer from soil to crop is very dependent on soil properties. These empirical models are generally based on a pH- dependent Freundlich equation. This equation was originally used to describe the solubility of metals in soils, and can be extended to fit the relationship between soil and plant metal concentrations [Eq. 1] (Brus et al., 2002; Brus and Jansen, 2004):

B Qplant = AQsoil [1] where Qplant and Qsoil are the metal concentrations in the crop and the soil, respectively, and A and B are the Freundlich coefficients. A is assumed to be a log-linear function of soil properties such as pH, soil organic matter (SOM) and clay content, so that Eq. [1] can be converted to Eq. [2]: log(Qplant) = A0 + A1 pH + A2 log(SOM) + A3 log(clay) + B log(Qsoil) [2]

The coefficients in Eq. [2], A0-3 and B, are obtained by multiple regression using paired soil- plant dataset. In some cases, the coefficients for SOM or clay are not significant in the regression model and, therefore, these terms are excluded from the model. In the derivation of a for Cd using the CLEA model, the relationship of log(CF) to soil pH was used, CF being the plant/soil concentration factor (Defra and EA, 2002a); this model is

Page 12 the same as Eq. [2] with the omissions of the SOM and clay terms and the coefficient B being fixed at 1.Table 5 shows some examples of empirical models for Cd.

The models presented in Table 5 were used to predict the potential for Cd accumulation in different crops, grass or vegetables across England and Wales, using the soil data from the National Soil Inventory (NSI) for England and Wales (McGrath and Loveland, 1992). In the NSI database, total soil Cd was determined by flame atomic absorption, which had a high detection limit (0.2 mg Cd/kg soil). A recent re-analysis of the NSI soils using a more sensitive method (X-RF) showed lower Cd concentrations, with the median being 0.36 of the old value. This factor was used to correct for the over-estimation of soil Cd in the NSI dataset. Table 5 shows the range and the median of the predicted Cd concentrations in the crops, if they were grown in all soils (n = 5577) across the country, and the percentage of the predicted concentrations that could exceed the respective limits. The cumulative frequency distributions of selected crops (wheat, barley and leafy vegetables) are presented in Figures 1 and 2.

These calculations show that exceedance of the Cd limit would be rare (0 – 1.6%) for wheat grain, barley grain, cabbage and potato grown in England and Wales. For wheat and barley grain, the predictions are in agreement with the nation-wide survey data showing few samples exceeding their respective limits (Adams et al., 2004; Chaudri et al., 1995), thus giving confidence to the model predictions. In contrast, leafy and root vegetables (including lettuce and carrot) have the potential to be more problematic, with the prediction that 24-53% of potential sites could lead to Cd concentrations that exceeded the limits, if these crops were grown on them. However, it is not known what proportion of these sites actually have these vegetables grown on them. Whilst there are no national survey data for these vegetable crops to back up the model predictions, it is known that these plants do tend to accumulate more Cd in their edible parts (Alexander et al., 2006; Kuboi et al., 1986). High Cd concentrations were also reported by Moir and Thornton (1989) in some vegetables grown in urban allotments. An important reason causing Cd exceedance in the model outputs is low measured soil pH (suggesting that many of these sites are not being maintained at pH ≥6.0 for agricultural use). If soil pH is maintained at ≥6.0, as would be expected in normal agricultural practices, exceedance of the Cd limits can be decreased markedly to ≤1.5% for most vegetable crops (Fig. 3). The only exception is root vegetables assessed with plant transfer function used in the CLEA model, which still shows 39% of sites with exceedance, examples might be if these crops were grown on allotments on soils with high levels of Cd. The model of Hough et al. (2004) for carrot, however, predicts that the exceedance would decrease from 26% to 0.4% if the soil pH is maintained at ≥6.0, in line with other vegetables.

Brus et al. (2002) used a soil to grass transfer model coupled with a grass to cattle transfer model (Table 5) to predict Cd concentrations in cattle kidney in the Netherlands. When these models are applied to the soils in England and Wales, only 0.2% is predicted to result in the kidney Cd concentration exceeding the 1.0 mg/kg fresh weight limit (Annex 2). Also, no scenario is predicted to exceed the feed Cd limit of 1.1 mg/kg dry weight (Annex 3).

It should be pointed out that extrapolation of the empirical models beyond the range of soil types or properties that were used to construct the model may not be valid. The uncertainty associated with model calculations is also difficult to assess. Nevertheless, the model calculations described above appear to be consistent in revealing that the potential issue of specific concern lies in the Cd accumulation by vegetable crops if they are grown in soils with elevated Cd, whereas for cereals and grassland the likelihood of exceeding the current limits in broad scale agriculture is very small.

Page 13

In light of the recent EFSA opinion on the need to reduce Cd exposures (ESFA (European Food Safety Authority), 2009a), there are proposals to reduce the Cd limits in offal and cereal (wheat) by half (Food Standards Agency, 2009a). Using this scenario, 28% of wheat and 3% cattle kidney are predicted to exceed the limits, if all soils in the NSI survey were used to grow wheat or grass for cattle and without adjustment of soil pH. This is likely to be an overestimate because not all soils are used in this way and the pH of the soils would be adjusted to achieve optimum crop growth. Refining this estimate would require data for all the sites, which was not available.

In contrast to Cd which is taken up by plant roots, the main entry route of Pb to plants is via atmospheric deposition (Dalenberg and Vandriel, 1990; Tjell et al., 1979). For this reason, no correlations were found between Pb concentration in wheat or barley grain and soil properties (Zhao et al., 2004). An empirical model for predicting Pb concentration in wheat grain was presented by Brus et al. (2004), although it explained only 36% of the variance in the grain Pb concentration. When this model is applied to the NSI data, the predicted wheat grain Pb ranges from 0.01 to 10.4 mg/kg dry weight, with 65% exceeding the 0.235 mg/kg DW limit (EC Food Regulation, 2006). The high percentage of exceedance is not supported by the nationwide survey data (Zhao et al., 2004), suggesting that applying the Pb model of Brus et al. (2004) to the soils in England and Wales is not valid. For vegetable crops, Hough et al. (2004) also presented empirical models for predicting Pb concentration using soil properties. Apart from the fact that these models are generally poor in accounting for the variance of the data, some models also contain negative coefficients for the soil Pb and positive coefficients for soil pH, meaning that plant Pb concentration would decrease with soil Pb concentration and increase with soil pH, which is unlikely to be correct. Based on the evidence available, it is concluded that the Pb concentrations of edible parts of food plants cannot be predicted reliably from soil measurements.

Page 14

Table 5. Empirical models for predicting Cd transfer from soil to crops or cattle and the applications of the models to the NSI database.

Predicted Cd in plant (mg/kg dry % sites that could exceed Cd weight) or kidney (mg/kg fresh limit if crops grown at all weight) Cd limit (mg/kg) sites Dry At fresh dry matter observed a Plant Model n Reference Minimum Maximum Median weight weight % pH At pH >6 Adams et al. Wheat grain log (Grain Cd) = 0.12 + 0.43 log(Soil Cd) – 0.16 pH 246 2004 0.011 0.46 0.08 0.2 0.225 88.8 1.6 0.14 log (Grain Cd) = 0.086 + 0.542 log(Soil Cd) – 0.126 Brus & Jansen Wheat grain pH – 0.218 log(SOM) 102 2004 0.01 0.67 0.067 0.2 0.225 88.8 0.3 0.2 Adams et al. Barley grain log (Grain Cd) = 0.20 + 0.82 log(Soil Cd) – 0.24 pH 84 2004 0.001 0.46 0.018 0.1 0.113 88.8 1.2 0.18 log (Grass Cd) = 0.168 + 0.489log(Soil Cd) – 0.125 Brus et al. Grass pH – 0.283 log(SOM) 115 2002 0.012 0.77 0.077

Grass to cattle Kidney Cd = (Cdgrass x Mgrass x fgrass + Cdsoil x Msoil x Brus et al. b kidney transfer fsoil)/(Mgrass + Msoil) ‐ 2002 0.037 3.64 0.27 1.0 0.2 0.2 Defra/EA Leafy vegetables Ln(Plant Cd) = 11.206 + Ln(Soil Cd) ‐1.634 pH ‐ 2002 0.001 351 1.06 0.2 3.03 6.6 33.5 1.5 Defra/EA Root vegetables Ln(Plant Cd) = 11.174 + Ln(Soil Cd) ‐1.646 pH ‐ 2002 0.001 325 0.95 0.1 0.667 15 53.2 39.4 log(Plant Cd) = ‐0.0023 + 0.294 log(Cd on SOM) – Hough et al. c Broccoli 0.153 pH 84 2004 0.038 0.804 0.20 0.05 0.333 15 9 0.4 log(Plant Cd) = ‐0.611 + 0.120 log(Cd on SOM) – Hough et al. Cabbage 0.623 pH 107 2004 0.000 0.003 0.0001 0.05 0.333 15 0 0 log(Plant Cd) = 1.24 + 0.355 log(Cd on SOM) – Hough et al. Carrot 0.311 pH 126 2004 0.034 4.11 0.49 0.1 0.847 11.8 26.3 0.4 log(Plant Cd) = 2.02 + 0.318 log(Cd on SOM) – Hough et al. Lettuce 0.359 pH 88 2004 0.072 14.2 1.43 0.2 3.03 6.6 23.9 0.02 a Dry matter % are from de Meeûs (2002). b Mgrass , amount of grass consumed by cattle = 15 kg/day; Msoil, amount of soil consumed by cattle = 0.5 kg/day; fgrass and fsoil, the transfer coefficient of Cd from grass or soil to animal kidney, respectively, both = 2.99. c Cd on SOM, mg Cd per kg soil organic matter.

Page 15

100

80

60 Wheat grain Barley grain

40 Cd limit for barley 20 Cd limit

Cumulative frequency (%) frequency Cumulative for wheat

0 0.0 0.1 0.2 0.3 0.4 0.5 -1 Predicted Cd concentration (mg kg dry weight)

Figure 1. Frequency distribution of predicted Cd concentrations in wheat and barley grains, using the models of Adams et al. (2004). Refer to Annex 2 for limits.

100

80

60 Leafy vegetables Leafy vegetables, soil pH > 6

40

20 Cd limit Cumulative frequency (%) frequency Cumulative

0 0 5 10 15 20 25 30 Predicted Cd concentration (mg kg-1 dry weight)

Figure 2. Frequency distribution of predicted Cd concentrations in leafy vegetables, using Defra-EA model. Refer to Annex 1 for limits.

For residents with allotments, the CLEA model estimated that consumption of home- grown vegetables and ingestion of soils attached to vegetables contribute 22% and 7%, respectively of the mean exposure to Pb (Defra and EA, 2002b). However, the Defra/EA soil guideline value for Pb is set based on the soil/dust ingestion pathway only, without consideration of the soil – plant – human pathway (Department of Environment Food and Rural Affairs and The Environment Agency, 2002b).

Page 16

No empirical models for As accumulation in food crops have been reported to date, but may be needed as a result of the EFSA opinion to reduce As intake by humans from food (European Food Safety Authority, 2009b). In the derivation of the soil guideline value for inorganic As, the Environment Agency used fixed soil-to-plant transfer factors for various vegetables and fruits, implying a linear relationship between plant and soil As concentrations (Environment Agency, 2009c). Williams et al. (2007) showed a log-log linear relationship, with considerable scatter, between soil As and grain As for wheat and barley in the UK. This is an area of uncertainty.

Little knowledge exists on interactions between contaminants in terms of uptake and accumulation within plants and animals and, therefore, food. It is known for example that Zn and Cd act in many cases as analogues and that there is competition for uptake, with Zn almost always present in much higher concentrations. So, when assessing the impact of Cd, Zn availability needs to be considered. However, models for even this well-known interaction are not yet available in the literature.

3. Organic contaminants

3.1 Introduction

There are a large and diverse range of organic contaminants (OCs) that may be present in food as a result of human activity or from natural sources. They may be introduced at any point in the food chain from growth and harvesting through to storage and processing. They may also result from environmental contamination. Exposure to these contaminants is potentially harmful in excess.

In this review, compounds or groups of compounds that are known to be present in soils or in materials applied to soils were identified, based on findings from Defra project SP0547 (Sources and Impacts of Past, Current and Future Contamination of Soil - Annex 2 : Organic Contaminants; Nicholson et al., 2008). These included well known persistent organic contaminants such as dioxins, polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAHs), as well as chemicals of emerging concern such as the brominated flame retardants and polyfluorinated alkyl substances. Background information on the uses, sources and routes of entry to soil of these and other OCs is provided in the final report for Defra project SP0547.

Some OCs may not be present in significant amounts or persistent in the soil environment. For others, the consumption of foods of terrestrial origin is not considered to be a potentially important route of human exposure or there is at present a lack of evidence on the significance of this exposure route. For these reasons, the following OCs or groups of OCs were not investigated in this review: phthalates, linear alkylbenzene sulfonates, oligochlorobenzyltoluenes (Ugilec), volatile organic compounds, human pharmaceuticals, veterinary medicines and biocides.

3.2. OC limits in food and soils

3.2.1 Background

International scientific committees such as the Joint FAO/WHO Expert Committee on Food Additives (JECFA) and the Joint FAO/WHO Meeting on Pesticide Residues (JMPR), regional scientific committees such as those of the European Union, and national regulatory agencies generally use a safety factor approach for establishing acceptable or tolerable intakes of substances that exhibit thresholds of toxicity. Page 17

Tolerable daily (TDI) or weekly (TWI) intakes are applied to contaminants in food and drinking water and are an estimate of the amount of a substance that can be ingested daily (or weekly) over a lifetime without a significant health risk. JECFA uses the term PTWI (provisional tolerable weekly intake) for contaminants that may accumulate in the body.

In this review, an assessment was made of those OCs which are known to pose a human exposure risk through the consumption of food products. OCs where a TDI has been established and/or those for which maximum residue limits (MRLs) or maximum permitted concentrations (MPCs) in foods have been set in EU or UK legislation were selected for further investigation. This approach was adopted to provide a benchmark against which to assess whether ‘significant’ (in terms of human health) concentrations in food were being approached. If no TDI or MPC for an OC has been set, then it is more difficult to determine the level at which the concentration of the OC in a food product might be considered ‘significant’ or ‘of concern’.

3.2.2 Limits in food or diet

There are limits for food (or animal feeding stuff) in place, or being considered, for the following OCs, for which further investigation of their behaviour in soils and uptake into plants was undertaken:

Dioxins and dioxin-like PCBs. The WHO introduced a revised TDI for dioxins of 1-4 pg WHO-TEQ/kg body weight (BW) in 1998 (WHO, 1998). In 2001, the European Commission and the Scientific Committee for Food proposed a TWI of 14 pg WHO- TEQ/kg BW for dioxins and dioxin-like PCBs (EC, 2001a). However in 2001, the UK, the Committee on Toxicity (COT) recommended a TDI of 2 pg WHO-TEQ/kg BW based on effects on the developing male reproductive system, which was considered adequate to protect against other possible effects (e.g. cancer and cardiovascular effects) and which is still used today (COT, 2001).

In addition, MPCs in food were specified in Commission Regulation 2375/2001 (EC, 2001b), amending Commission Regulation 466/2001 (EC, 2001c) and updated in Commission Regulation 1881/2006 (EC, 2006), enforced by The Contaminants in Food (England) Regulations (SI, 2009). Around 80% of human exposure to dioxins and dioxin-like PCBs is from consumption of animal products. Hence, the limits set in EC (2006) are mainly for foods derived from animals (e.g. meat, liver, milk and dairy products, eggs, animal fats), although there is also a limit for vegetable oils (Table 6). Note: levels of these contaminants are currently being reviewed and may be reduced (FSA, 2009a). There are no limits for dioxins and dioxin-like PCBs in cereals, fruit, vegetables or salads. Dioxin levels in animal feeding stuffs are regulated in the UK (SI, 2005), where the limit in all feeding materials of plant origin is 0.75 ng WHO- PCDD/F-TEQ/kg (Table 7).

• Non-dioxin like PCBs. The Joint FAO/WHO Expert Committee on Food Additives concluded that, owing to the limitations of the available data, it was impossible to establish a precise numerical value for a tolerable intake of total PCBs for humans (WHO, 1990). A similar conclusion was reached by EFSA in 2005 (EFSA, 2005). However, several national governments are employing TDIs for PCBs for risk management purposes. For example, in Germany a TDI for PCBs of 1–3 μg/kg BW has been suggested (WHO, 2000). Indeed, the European Commission is proposing the introduction of limits for non dioxin-like PCBs in animal products and infant feeds (Table 8). PCBs are not currently regulated in livestock feeding stuffs. Page 18

• PAHs. Benzo(a)pyrene (BaP) is used as a marker for the occurrence and effect of carcinogenic PAHs in food (EC, 2006). However, a recent EFSA opinion (EFSA, 2008), based on newer data, suggested that BaP alone is not a suitable indicator for PAH contamination and has recommended that new PAH limits be based on BaP together with three other PAHs: benz(a)anthracene, benzo(b)fluoranthene and chrysene. Although food constitutes the main route of intake of PAHs, the Joint FAO/WHO Expert Committee on Food Additives was unable to establish a tolerable intake for BaP (WHO, 2003).Maximum levels of BaP have been specified in certain foods containing fats and oils, although the major routes of contamination are smoking or drying processes, or environmental pollution, rather than the soil-plant uptake route. MPCs in infant milk and food are also specified (Table 9).

• Pentachlorophenol (PCP). PCP is a fungicide used mainly used in wood preservatives and has not been approved for food use. It is also of concern as it contains trace levels of chlorinated dioxins and furans created as unintentional by- products of the manufacturing process. The Codex/JMPR have not specified a TDI or MRLs for PCP, neither have food limits been set at the EU level except for guar gum from India. However, some individual countries have specified limits for PCP in certain foodstuffs (Table 10).

• Hexachlorobenzene (HCB). The WHO Task Group on Environmental Health Criteria for HCB derived a TDI of 0.17 μg/kg BW per day for non-neoplastic effects and a health based guidance value of 0.16 μg/kg BW per day for neoplastic effects (WHO, 1997). There is no EU legislation on maximum HCB levels in food.

• Chlorinated paraffins (CPs). TDIs for different types of CPs were derived by the World Health Organization (WHO) in its 1996 review (WHO, 1996). For short and medium chain CPs a TDI of 100 μg/kg BW/day was derived. In addition, WHO recommended that daily doses of short chain CPs for the general population should not exceed 11 μg/kg BW for neoplastic effects (WHO, 1996).

• Brominated flame retardants (BFRs). The European Commission has repeated its previous informal call for data on BFR levels in food (FSA, 2009c), but at present there are no limits in food products.

• Nonylphenol (NP). A TDI of 5 μg/kg BW has been proposed by the Danish Institute of Safety and Toxicology (Nielsen et al., 2000), but this has not yet been adopted at the European level.

• Perfluoroalkyl substances (PFAs). TDIs have recently been set by the European Food Safety Authority of 0.15 and 1.5 μg/kg BW for perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA), respectively (FSA, 2009c).

. MRLs for pesticides are specified by the Codex Committee on Pesticide Residues for a wide range of pesticides in a very large number of food products including cereals, animal products, fruit and vegetables (FAO/WHO, 2009). Pesticide behaviour in soils and uptake in plants was not reviewed further in this study (except where the substance was included in one of the above contaminant groups).

Page 19

Table 6. Summary of dioxin and dioxin-like PCB limits in human food or diets of terrestrial origin.

Commodity Maximum levelsa

Meat and meat products of : - bovine animals and sheep 4.5 - poultry 4.0 - pigs 1.5 Liver of animals above 12.0 Raw milk and dairy products 6.0

Hens eggs and egg products 6.0 Fat of the following animals:

- bovine animals and sheep 4.5 - poultry 4.0 - pigs 1.5 Mixed animal fats 3.0 Vegetable oils and fats 1.5

a EC (2006)

Table 7. Summary of dioxin limits in animal feeding stuffs of terrestrial origin.

Product intended for animal feed Maximum level (ng WHO-PCDD/F- TEQ/kg)a All feeding materials of plant origin 0.75 including vegetable oils and by-products Animal fat 2.0 Other land animal products 0.75 Compound feeding stuffs 0.75 aSI (2005). Assumes moisture content of 12%.

Page 20

Table 8. Summary of currently proposed non dioxin-like PCB limits in human food of terrestrial origin.

Commodity Proposed maximum levels in fooda

Meat and meat products of : - bovine animals and sheep 40 ng/g fat - poultry 40 ng/g fat - pigs 40 ng/g fat Liver of animals above 40 ng/g fat Raw milk and dairy products 40 ng/g fat Hens eggs and egg products 40 ng/g fat Fat of the following animals: - bovine animals and sheep 40 ng/g fat - poultry 40 ng/g fat - pigs 45 ng/g fat Mixed animal fats 40 ng/g fat Vegetable oils and fats 40 ng/g fat Foods for infants and young children 1.0 ng/g fresh weight aFSA (2010). Sum of six PCBs: 28, 52, 101, 138, 153, 180.

Table 9. Summary of BaP limits in human food of terrestrial origin.

Maximum levels in food (μg/kg wet Commodity weight)a

Oils and fats 2.0 Smoked meat and smoked meat products 5.0 Processed cereal based foods and baby food 1.0 Infant formula and follow on milk 1.0 Dietary foods for infants 1.0 aEC (2006).

Page 21

Table 10. Summary of limits in human food for PCP.

Country Maximum Residue Limits (mg/kg)

Belgium Mushrooms 0.05; other 0a (0.01).

Germany All foods of plant origin 0.01. Israel Mushrooms and other foods 0.05. Netherlands Mushrooms 0.05; other 0 (0.01)b.

Switzerland Milk 0.05. Former Yugoslavia Unnamed commodities 0.01. EU Guar gum from India (0.01)c Source: FAO/UNEP (1996). aISO designation. bResidues shall be absent. The highest concentration at which this requirement is still deemed to have been met is given in parentheses. cCommission Regulation 258/2010.

3.2.3. Limits in soils

In the UK, there are no specified ‘safe’ limits for OC concentrations in agricultural soils (or soil amendments such as sewage sludge and compost) either to protect human health or the environment. Quality standards for various OCs were proposed in the Working Document on Sludge (EC, 2000) and have been established in a number of European and other countries. Following a detailed risk assessment, the US EPA concluded that the regulation of OCs was unnecessary to protect human health and the environment where sewage sludges were applied to soil (US EPA, 2003). Researchers in the UK have also argued that there is no scientific justification for setting limits on OCs in sewage sludges (e.g. Jones and Stevens, 2002; Stevens and Jones, 2003; Stevens et al., 2003; Rhind et al., 2005a,b).

However, a set of preliminary, health related numerical limits for OCs in soil was developed by Chang et al. (2002) for land application of wastewater and sewage sludge (Table 11). The limits were developed considering only the food chain transfer of contaminants via the ‘waste’ → soil → plant → human route, based on potential OC intakes from consumption of grain, vegetables, root/tuber, and fruit on the basis that food chain transfer is the primary route of human exposure to environmental contaminants. Daily intakes of grain/cereal, vegetables, root/tuber and fruit were assumed to account for about 75% of daily adult food consumption. The exposure scenario assumed that most exposed individuals were the adult residents (60 kg in body weight) of a land application area whose entire consumption of grain, vegetables, root/tuber and fruit were produced from the affected soils (i.e. a worst case scenario). Their daily intake of OCs from consumption of grain, vegetables, root/tuber and fruits was assumed to account for 50% of the ADI, with the remaining 50% credited to the background exposure.

Page 22

Table 11. Preliminary health-related guideline maximum permissible concentrations of selected OCs in soils receiving sewage sludge and/or untreated municipal wastewater (Chang et al., 2002).

Organic compound Soil concentration (mg/kg)

PAH (as BaP) 16 HCB 1.40 PCP 14 PCBs 0.89 Dioxins 0.00012

3.3. Behaviour of OCs in soil and plant uptake

3.3.1 Introduction

The behaviour of OCs in soil depends on their physical and chemical properties and on the physical, chemical and biological nature of the soil. Key properties in determining the environmental fate and behaviour of OCs include polarity, vapour pressure and chemical stability.

Polarity. In general, the more polar an OC, the higher its water solubility and vice versa. OCs of low polarity tend to be lipophilic (i.e. soluble in fats/oils and non-polar solvents) and of low water solubility. The balance between the lipophilicity and hydrophilicity of an OC is indicated by its octanol-water partition coefficient (KOW = concentration in octanol/concentration in water). KOW values for highly lipophilic OCs are very large and hence are commonly expressed as log values to the base 10.

Log KOW has been used extensively to predict OC environmental fate. In general, OCs with high KOW values have a greater tendency to adsorb onto soil particles and organic matter, and tend to be less readily broken down by biological processes. Hence, KOW values can be used as an indication of environmental persistence and can potentially be used to identify those which may merit more targeted investigation (IC Consultants, 2001).

The organic carbon-water partition coefficient (Koc) is the ratio between the concentration of an OC in organic carbon and in water. In general, OCs with higher Koc values tend to adsorb more strongly onto organic carbon, whilst those with lower Koc, values are more readily leached. Thus, Koc values, which are usually expressed on a log scale, indicate the likelihood of leaching through soil.

Vapour pressure. High vapour pressure values indicate that an OC will readily volatilise. As a general guide, compounds with a Henry’s Law constant >10-3 atm (mol-1 m-3) can be lost by volatilisation. The significance of volatilisation losses of specific OCs can be predicted using Henry’s constant and Kow (IC Consultants, 2001).

Stability. OCs which are highly resistant to breakdown are described as refractory or recalcitrant (Walker, 2001). Many are not photochemically stable and may be Page 23 oxidised or undergo molecular rearrangement when exposed to sunlight. Some are susceptible to hydrolysis especially under conditions of high pH, although highly halogenated OCs such as some PCBs and dioxins tend to be resistant to oxidation and other mechanisms of chemical degradation. The presence of charged groups will also influence the behaviour of an OC in the environment. Some OCs exist as anions or cations in solution which may become bound to soil organic macromolecules or minerals of the opposite charge (Walker, 2001). OC persistence, which is usually based on a measured half-life value determined by controlled incubation in soil, is often used to assess its potential impact on the environment.

Soil properties. Heavy soils, high in organic matter and clay, adsorb lipophilic OCs (i.e. those with a high KOW) that would be more mobile or bioavailable in light, sandy, low organic matter soils. In contrast, hydrophilic or polar OCs are less adsorbed to soil colloids and are consequently more rapidly degraded by soil enzymes and more readily leached.

Plant uptake. Wild and Jones (1992) identified the principal pathways by which OCs in the soil can enter a plant:

• root uptake and subsequent translocation; • foliar uptake of vapour from the surrounding air; • uptake by external contamination of shoots by soil and dust, followed by retention in the cuticle or penetration through it; and • uptake and to lipid rich parts of some plants (e.g. carrot skin).

The main focus of this review is on the first pathway (OC uptake via the roots), which is affected by OC physico-chemical properties, environmental conditions, soil type and plant characteristics. Plant uptake can be described as a series of consecutive partition reactions, with partitioning occurring between soil solids and soil water, soil water and plant roots, plant roots and the transpiration stream, and the transpiration stream and plant stem. This partitioning can be related to the KOW, such that compounds with high KOW values (i.e. lipophilic OCs such as dioxins, PAHs and PCBs) are most likely to be sorbed by the soil and/or plant root. Chemicals with lower KOW values are more likely to be translocated within the plant and may reach the above ground portions of the plant (Wild and Jones, 1992). Less research has been undertaken on the effect of soil properties on plant OC uptake, although Topp et al. (1986) studied the uptake from soil of 14C from various 14C-labeled OCs by barley and cress seedlings and found a negative correlation between barley root OC uptake and soil organic carbon.

A more detailed discussion of the research on specific groups of OCs is given below.

3.3.2 Dioxins.

Dioxins are generally strongly sorbed and persistent in soil. Due to their low vapour pressure and water solubility and their strong sorption to , dioxins are generally immobile in soils and sediments (ATSDR, 1998). The half-lives of different congeners have been reported to range from c.10 years (di Domenico et al. 1980) to c.150 years (Sinkkonen and Paasivirta, 2000). Somewhat surprisingly, the UK Soil and Herbage Survey reported a drop in soil dioxin concentrations (for both rural and urban soils), despite the reported persistence of dioxins in soils, which may indicate that estimates of dioxin half-lives in soil are too high (Creaser et al., 2007a), or that the significant reductions in emissions have reduced concentrations at or near the surface of the soil. Page 24

Although dietary intake represents over 90% of human exposure to dioxins, this is mainly from the consumption of meat, dairy products and fish, with intake from vegetables or fruits considered to be negligible. Nevertheless, plants are the first link in the food chain and the significance of plant uptake as a dietary source of dioxins has been extensively researched (Hülster et al., 1994). Various experiments have shown that root uptake of dioxins is minimal. For example, Engwall and Hjelm (2000) found that dioxin concentrations in carrot roots were independent of the amounts applied in sewage sludge, and Hülster et al. (1994) reported that plant shoot concentrations were very low even in highly contaminated soils. Despite the general consensus that there is negligible uptake of dioxins by crop plants and that atmospheric sources are mainly responsible for crop contamination, there is some evidence indicating that for certain vegetables (e.g. courgette and pumpkin), root uptake may be the dominant pathway (Hülster et al., 1994).

3.3.3 Polychlorinated biphenyls (PCBs)

Like dioxins, PCBs are highly sorbed and persistent in soil as a result of their low water solubility and high KOW. Once deposited on soil, PCBs may be adsorbed by organic matter, degraded by micro-organisms or photodegradation, or re-emitted by volatilisation.

Adsorption by organic matter is an important retention mechanism for PCBs in soil, with Meijer et al. (2003) reporting a correlation between soil PCB levels and soil organic matter. However, Doick et al. (2005) concluded that sequestration of PCBs by the mineral fraction of the soil may be more important than previously thought. There is evidence to suggest the degradation of lower molecular weight congeners is dominated by aerobic microbial activity, while the de-chlorination of higher molecular weight PCBs to lower molecular weight congeners is predominantly due to anaerobic activity (e.g. Haluska et al., 1995). In some cases, particularly for the lighter congeners, volatilisation is the main mechanism by which PCBs are lost from soil (Beyer et al., 2000).

Estimates of the half-life of PCB congeners in soil vary significantly although, in some cases, not all loss mechanisms were included in the studies. The half-lives of freshly incorporated PCB congeners in a range of soils varied from 90 to 98 days for congener 28, and from 5,776 to 17,239 days for congener 180 (Ayris and Harrad, 1999), with the main loss mechanism inferred to be volatilisation. Sinkkonen and Paasivirta (2000) suggested half-lives for PCBs in soil ranging from 1,083 days for congener 28, to 13,750 days for congener 180, but only photodegradation in the upper soil layers was considered. Doick et al. (2005) reported half-lives for PCB 28 and 52 of c.4000 and c.4100 days, respectively, which were considerably longer than those previously estimated.

Calculations based on the fugacity approach (fugacity is a measure of the ‘tendency to escape an environmental compartment’) have suggested that UK soils may now be net sources of lower PCB congeners to the atmosphere, but still long-term sinks for the heavier congeners (Cousins and Jones, 1998). However, a more recent study – also using a fugacity approach – concluded that UK soils are still net sinks of PCBs 28, 153 and 180 (Dalla Valle et al. 2005). Studies on archived soil samples from Rothamsted Experimental Station (Hertfordshire) indicated that, alongside significant reductions in the total PCB burden in soil, the contribution of the lighter congeners to total PCB concentrations had declined while the heavier congeners had increased (Lead et al., 1997). Re-emission and participation in the so-called ‘grasshopper effect’ (Gouin et al., 2004), in which the molecule undergoes a series of emission– Page 25 deposition cycles, is likely to be more important for the lower molecular weight congeners.

Like dioxins, the uptake of soil PCBs by plants is generally thought to be negligible because of the strong sorption to soil organic matter and clay (ATSDR, 2000), with levels of PCBs in field-grown plants largely reflecting deposition from the atmosphere (Creaser et al., 2007b). As a result, below-ground plant parts (such as potatoes) accumulate the lowest PCB levels although greater uptake from the soil can occur in certain root crops (e.g. carrots) as a result of the partitioning of PCBs into the lipid- rich skin (ATSDR, 2000).

The influence of soil properties on PCB uptake has not been widely investigated, however, O’Connor et al. (1990) found that crop (fescue, lettuce and carrot) PCB concentrations from a clay loam soil were lower than from a sandy loam soil, which they attributed to reduced PCB transfer to the crop due to increased sorption on clay particles. Soil to plant concentration factors (CF – the concentration quotient of plant to soil) of soil PCBs are estimated to be <0.02 for most plant species (ATSDR, 2000).

3.3.4 Polycyclic aromatic hydrocarbons (PAHs)

Most PAHs reach agricultural soils via aerial deposition. Like dioxins and PCBs, many PAHs are highly sorbed and persistent in soil, although potential loss routes include volatilisation and microbial degradation (Yang et al., 2009). PAH sorption to soil has been shown to be enhanced by the application of sewage sludge (Wild et al., 1995). Calculations on the fugacity of PAHs and experimental studies suggest that soils could now be a source of lower molecular weight PAHs (i.e. the more volatile) to the atmosphere but still act as a long-term sink for the heavier PAHs (Cousins and Jones, 1998). However, measurement of the depth distribution of PAHs in soils has indicated no depletion near the soil surface suggesting that, in practice, volatilisation is not a significant loss process (Cousins et al., 1999). In most cases, microbial biodegradation is the main removal process for PAHs in soil (Wild et al., 1991). There is considerable variation in the published literature on PAH half lives in soils (Table 12).

Page 26

Table 12. Estimated half lives of some PAHs in soils derived from experimental studies and suggested half-lives derived from model calculations.

Compound Estimated half- Source Modelled life half-lifea

Naphthalene 2.1–2.2 daysPark et al. (1990) Anthracene 10–134 daysPark et al. (1990) 7.9 years Wild et al. (1991) 0.34–1.15 Phenanthrene 16–35 daysPark et al. (1990) 5.7 years Wild et al. (1991) 0.34–1.15 Fluoranthene 268–377 daysPark et al. (1990) 7.8 years Wild et al. (1991) 1.15–3.4 Pyrene 199–260 daysPark et al. (1990) 8.5 years Wild et al. (1991) 1.15–3.4 Chrysene 371–387 daysPark et al. (1990) 8.1 years Wild et al. (1991) 1.15–3.4 Benzo(a)pyrene 229–309 daysPark et al. (1990) 8.2 years Wild et al. (1991) 1.15–3.4 3.2 years Doick et al. (2005) Dibenz(a,h)anthracene 361–420 daysPark et al. (1990) 1.15–3.4 Acenapthylene - 0.11–0.34 Acenapthene >3.2 years Wild et al. (1991) 0.11–0.34 Fluorene >3.2 years Wild et al. (1991) 0.34–1.15 Benz(a)anthracene 8.1 years Wild et al. (1991) 1.15–3.4 Benzo(b)fluoranthene 9.0 years Wild et al. (1991) 1.15–3.4 Benzo(ghi)perylene 9.1 years Wild et al. (1991) 1.15–3.4 aMackay et al. (1992).

Gaseous deposition is generally accepted as the principal pathway for PAH accumulation in vegetable crops (Kipopoulou et al., 1999), with PAH movement to plants via root uptake not believed to be a major pathway, because PAHs are strongly retained by soil organic matter (Wild and Jones, 1992). Duarte-Davidson and Jones (1996) studied the fate of PAHs in sewage sludge amended agricultural soil and suggested that PAHs may strongly adhere to root surfaces, but the potential for root uptake and translocation was low. Similarly, Wild et al. (1991) and Simonich and Hites (1995) indicated that lipophilic OCs, including PAHs, partition to the root skin or to soil particles and are not drawn into the inner root, because this part of the translocation system was water based. Carrots represent a potential ‘worst case’ for absorption of lipophilic compounds into the skin due to their high lipid content, although the skin is usually removed prior to consumption. However, uptake of a small amount of PAHs by vegetable roots from soil cannot be totally ruled out. For example, Kipopoulou et al. (1999) detected PAHs in peeled carrot roots and suggested that the (more bioavailable) low molecular weight PAHs may move through into the core.

Several authors have measured concentration factors for PAHs. For example, Tao et al. (2004) quoted an average CF for all vegetables studied of 0.029, which fell within the range 0.001 to 0.4 reported elsewhere in the literature. Harms (1996) reported a CF for anthracene of 0.014 and for phenanthrene c.0.02. Similarly, Alcock et al. (2000) reported CFs ranging from 0.0001 to 0.33 for benzo-a-pyrene and from 0.001 to 0.18 for 17 other PAH compounds.

Page 27

3.3.5. Pentachlorophenol (PCP)

PCP may enter agricultural soils as a result of its use as a pesticide, with PCP released into the atmosphere and transported to soils via wet and dry deposition (WHO, 2003). PCP is less strongly bound to soil than dioxins, PCBs and PAHs, with reported half lives ranging from 10 days (Bellin and O’Connor, 1990) to 38-100 days (Wild and Jones, 1992).

PCP has a propensity to leach and contaminate surrounding environments. Whilst soil organic matter can enhance PCP adsorption, association with dissolved organic matter in soil solution can increase its water solubility and mobility (Scelza et al., 2008). PCP is sparingly soluble in water at soil pH values between 5 and 7. However, Mills et al. (2006) reported that it was more mobile under high pH conditions (≥7), where it is present primarily as the phenolate anion. It has strong biocidal activity making it resistant to microbial degradation, although in wet soil, photolysis may be a significant route of PCP loss (Donaldson and Miller, 1997).

There is very limited information available on the uptake of PCP by crop plants. Casterline et al. (1985) grew soybean and spinach in a mixture of soil and sand containing 10 mg/kg PCP and detected PCP and its metabolites in the harvested plants. Bellin and O’Connor (1990) studied the effects of sludge applications on plant (fescue, lettuce, and carrot) PCP uptake and detected only very low PCP concentrations in fescue and lettuce, and none in carrot tissues, apparently as a result of rapid PCP degradation in the soil limiting PCP availability to the plant. CFs for PCP were <0.01 for all crops, again suggesting little plant PCP uptake.

3.3.6 Hexachlorobenzene (HCB)

HCB (formerly used as a fungicide, now a banned substance) is widely distributed in the environment because of its resistance to degradation, although slow microbial biodegradation in the soil can occur (WHO, 2003). HCB can volatilize from the soil surface, but strongly adsorbs onto organic matter and is generally considered immobile with respect to leaching. Its half-life in soils is estimated to range from 3 to 6 years (WHO, 2003).

There is very limited data on HCB uptake by plants. Wang and Jones (1994) investigated the uptake of chlorobenzenes (CBs) by carrots grown in soil treated with different rates of sewage sludge or spiked with CB compounds. Sludge-amended soil tended to retain more CBs than spiked soil. Both carrot foliage and roots took up CBs from all the soil treatments. CB offtake in carrot foliage was affected by the yields of the above-ground parts, implying that the better growing leaves might take up CBs more effectively. CB offtake by roots increased with soil CB concentrations. Penetration of CBs through the carrot peel to the core was restricted and the resistance increased with molecular weight. CB uptake by carrots only accounted for a very small proportion of CBs applied to soil.

3.3.7 Chlorinated paraffins (CPs)

There is a very limited database on the environmental fate of short chain CPs (Alcock et al., 1999), although they are of low volatility and water solubility. Their high log Kow values indicate a strong propensity to bind to soil, and so they are not mobile and are relatively persistent. Biodegradation of short chain CPs has been shown to take place relatively slowly, with the biodegradation rate higher for grades with lower

Page 28 chlorine content and where bacteria had become acclimatised to CPs (Eurochlor, 2010).

Very little information is available on uptake by plants. However, a study investigating the accumulation of medium chain CPs in carrot roots was carried out using soils spiked to achieve concentrations that might be found in sewage sludge treated soils (Thompson et al., 2005). The mean measured soil concentration over the 70 day study period was 4.9 mg/kg dry weight and from this a mean BCF over days 50-70 of 0.045 was derived.

3.3.8 Nonylphenol (NP)

NP is generally regarded as being a persistent organic contaminant, with its high log Kow value suggesting it should be strongly bound to soil particles. However, there is contradictory information in the literature on the behaviour of NP in soils. For example, Roberts et al. (2006) reported that NP was not persistent in soil and could easily be leached, and Ghanem et al. (2006) also demonstrated that NP could be partly transferred in soil . In contrast, Soares et al. (2008) stated that NP mobility in soil was low, with almost 99% found within 30 cm of the soil surface 730 days after application.

Mineralisation has been shown by several authors to be a potential mechanism of NP loss from soils. Roberts et al. (2006) found that NP mineralization was concentration- dependent, and was stimulated by the addition of organic residues (e.g. sewage sludges, glucose, dead roots). Other studies have indicated that NP mineralization was a biotic process affected by temperature (e.g. Topp and Starratt, 2000). Volatilization of NP from soils is not significant, with only 0.22% removed over a period of 40 days (Soares et al., 2008).

There have been few investigations on NP uptake into plants, although its presence in foodstuffs would suggest that uptake can occur. Ghanem et al. (2006) studied the fate of sewage sludge spiked with 14C-labelled NP applied to the soil and found partial transfer to plant seedlings. However, Roberts et al. (2006) showed that the uptake of NP by plants was very low, and that NP had no negative effect on or plant growth unless present at extreme concentrations.

3.3.9 Perfluoroalkyl substances (PFAs)

There is very limited information on the behaviour of PFAs in soils and very few studies have been carried out on environmental solid matrices (Li et al., 2010). However, PFAs do not hydrolyze, photolyse or biodegrade under environmental conditions and are persistent in the environment (Giesy and Kannan, 2002; OECD, 2002).

There is very little published information on plant uptake of PFAs. However Stahl et al. (2009) found that plant (spring wheat, oats, potatoes, maize and ryegrass) PFOA/PFOS concentrations varied with soil concentrations, with PFOA values higher than PFOS values in all plants except potatoes. Uptake and storage were much greater in the vegetative portion of the plant than in the storage organs, particularly between the grain/ear and straw of spring wheat, oats and maize. However, transfer from soil to crops was a possible explanation for the presence of PFAs in foodstuffs and in human body fluids.

Page 29

3.3.10 Summary

Most information on OC behaviour in soils was found for the well known contaminants (i.e. dioxins, PCBs and PAHs), which are all strongly adsorbed by and persistent in soil. Of the other OCs investigated, most were highly persistent in soil and resistant to degradation, except for PCP and NP, where there was some evidence that they may be leached from or biodegraded in soil (NP only).

Plant uptake studies have also largely been confined to dioxins, PCBs and PAHs. OC uptake via roots and translocation to the above-ground portion of crop plants only occurs to a limited extent, even where relatively large quantities of OCs are present in the soil as a result, for example, of sewage sludge or compost applications (Smith, 1996; WRAP, 2009). There is evidence of slight enrichment of some OCs in certain root crops (e.g. carrots), but the transfers are very inefficient, with enrichment confined to the skin, which is normally removed before consumption.

Thus for most OCs, the major pathway into plants is likely to be either direct deposition from the atmosphere or contamination of plant surfaces with dust or soil particles (and possible subsequent translocation into the body of the plant). This implies that soil OC concentrations are of ‘indirect’ importance in terms of entry into food crops, although washing/peeling vegetables before eating will decrease the quantities ingested.

3.4 OC concentrations in soils

The following section summarises data on OC concentrations in agricultural soils in England and Wales used for growing food crops or for grazing livestock. Urban soil concentrations are reported for comparison where available, as some of these soils (i.e. the less contaminated soils) may be used for growing small quantities of food crops in gardens and allotments. There is no national database of soil OC concentrations equivalent to the National Soil Inventory (NSI) data for metals (McGrath and Loveland, 1992). The following main data sources were used:

• UK Soil and Herbage Survey (UKSHS) which reported dioxin, PCB and PAH concentrations in soil samples collected in 2001/2 from 203 rural, urban and industrial sites in England, Northern Ireland, Scotland and Wales (Creaser et al., 2007a;b;c); • Countryside Survey from the year 2000 (CS200) which reported PCB and PAH concentrations in 97 soil samples from England and Wales (Black et al., 2002).

Data from these sources, together with other published data from UK studies are summarised in Table 13. Direct comparison of measured concentrations between studies was not always possible because different numbers and combinations of congeners were measured. For example, the UKSHS reported total PAH concentrations for 22 congeners, whereas CS2000 reported data for 14 PAH congeners.

Page 30

Table 13. Concentrations of some OCs in agricultural, rural and urban soils from England and Wales.

Compound England Wales Comment Source Dioxins: PCDD/F (WHO-TEQ ng/kg) 2.53 2.02 Rural soils (median) UKSHS (Creaser et al., 2007a) PCDD/F (WHO-TEQ ng/kg) 11.1 3.47 Urban soils (median) UKSHS (Creaser et al., 2007a) PCDD/F (ng/kg) 279 (69–792) - 14 UK grassland soils (mean; range in brackets) Nam et al. (2008) PCDD/F (TEQ ng/kg) 0.2 - Mean value for UK soils EMEP (2009)

PCBs: ΣPCB (26 congeners) (μg/kg) 0.97 1.02 Rural soils(median) UKSHS (Creaser et al., 2007b) ΣPCB (26 congeners) (μg/kg) 2.52 1.88 Urban soils (median) UKSHS (Creaser et al., 2007a) ΣPCB (35 congeners) (μg/kg) 2.08 - Rural soils in England and Wales (median) CS2000 (Black et al., 2002) PCB (μg/kg) 1.23 (0.30–3.11) - 14 UK grassland soils (mean; range in brackets) Nam et al. (2008) PCB-153 (μg/kg) 0.05 Mean value for UK soils EMEP (2009) ΣPCB (20 congeners) (μg/kg) 32-57 2 control soils (1988/1990) Alcock et al. (1996) ΣPCB (20 congeners) (μg/kg) 22-106 8 sludged soils (1988/1990) Alcock et al. (1996)

PAHs: ΣPAH (22 PAHs) (μg/kg) 936 958 Rural soils (median) UKSHS (Creaser et al., 2007c) ΣPAH (22 PAHs) (μg/kg) 10,500 5,380 Urban soils (median) UKSHS (Creaser et al., 2007a) ΣPAH (14 PAHs) (μg/kg) 521 - Rural soils in England and Wales (median) CS2000 (Black et al., 2002) PAH (μg/kg) 704 (56–11,180) - 14 UK grassland soils (mean; range in brackets) Nam et al. (2008) BaP (μg/kg) 67.2 57.8 Rural soils (median) UKSHS (Creaser et al., 2007c) BaP (μg/kg) 43.7 - Rural soils in England and Wales (median) CS2000 (Black et al., 2002) BaP (μg/kg) 130 - Soil from Rothamsted Experimental Station (1980) WHO (1998) BaP (μg/kg) 0.04 - Mean value for UK soils EMEP (2009) Anthracene (μg/kg) 13 - Soil from Rothamsted Experimental Station (1980) WHO (1998) Napthalene (μg/kg) 23 - Soil from Rothamsted Experimental Station (1980) WHO (1998)

Chlorinated paraffins (mg/kg) <0.1 - Soils mostly from north west England Nicholls et al. (2001) - Nonylphenol (μg/kg) Not detected - 2 English soils (background) Gibson et al. (2005) 1053-4368 2 English soils (sludge amended) Gibson et al. (2005)

Page 31

Dioxins. The UKSHS (Creaser et al., 2007a) reported data on 17 polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs). Because of the persistence of dioxins, soil concentrations reflect emissions and atmospheric deposition over the last 10–30 years. The results indicate that, historically, land use was the main determinant of dioxin concentrations. Median concentrations in urban and industrial soils were approximately 2 to 3-fold those in rural soils, reflecting the presence of significant localised dioxin sources in urban and industrial areas (Table 13). Comparison with earlier surveys indicated that dioxin concentrations in soils were falling, reflecting significant reductions in dioxin emissions since the early 1990s. Mean dioxin concentrations in UK soils from the European Monitoring and Evaluation Programme (EMEP, 2009; Table 13) were substantially lower than those in the UKSHS, probably because areas with much lower soil dioxin concentrations (e.g. Scotland) were included in the mean.

PCBs. The UKSHS analysed 26 selected PCBs in soils and herbage (Creaser et al., 2007b). Median PCB concentrations in urban and industrial soils were approximately double those in rural locations, indicating that, historically, urban and industrial areas were significant sources of PCBs (Table 13). Alcock et al. (1996) reported that soil PCB concentrations declined over time (due to volatilisation losses), and that recent measurements (up to 20 years after sludge application) showed little difference in PCB concentrations between the control and sludge treated soils, although there were much greater differences where sludge had been freshly applied.

PAHs. The UKSHS analysed 22 PAHs in soils and herbage (Creaser et al., 2007c). Across the UK, concentrations of PAHs in urban and industrial soils were 5 to 8-fold higher than those in rural areas (Table 13). As for dioxins, mean BaP concentrations in UK soils from the European Monitoring and Evaluation Programme (EMEP, 2009; Table 13) were substantially lower than those in the UKSHS and CS2000.

Hexachlorobenzene. There was very limited data on HCB concentrations in UK agricultural soils. However, Nam et al. (2008) reported HCB concentrations of 190 pg/g (range 22–681 pg/g) for 14 UK grassland soils. Chlorinated paraffins. There were very few data on soil concentrations of chlorinated paraffins. However, Nicholls et al. (2001) reported polychlorinated alkane concentrations of <0.1 mg/kg dry soil, for samples collected predominantly in industrial areas of the north west of England where sewage sludge had been applied.

Nonylphenol. There was very little information on NP concentrations in UK soils. In a detailed review, Soares et al. (2008) reported elevated concentrations in soils exposed to high addition rates of sewage sludge (1,400-1,600 μg/kg) compared with unamended soils or soils fertilised with manure or limited amounts of sewage sludge (0.01–0.98 μg/kg), although these were not studies from the UK. Gibson et al. (2005) were unable to detect NP in two English soils, but found NP in sludge-amended soils at appreciable concentrations (>1,000 μg/kg).

Pentachlorophenol and PFAs. No data could be found on PCP or PFA concentrations in agricultural soils from England and Wales.

3.6 OC concentrations in plants

Table 14 presents data on OC concentrations in crops and herbage in England and Wales. Data sources used included the UK Soil and Herbage Survey (UKSHS) which reported dioxin, PCB and PAH concentrations in herbage samples collected in 2001/2 from 203 rural, urban and industrial sites in England and Wales (Creaser et Page 32 al., 2007a;b;c), together with other published data from UK studies. Total Diet Survey data from the Food Standards Agency are included for completeness, although as they are composites of c.100 samples intended to measure dietary exposure they are not a ‘true’ measure of typical levels in food.

Data from the UKSHS showed that median dioxin concentrations in herbage from England (1.02 and 0.9 WHO-TEQ ng/kg dry weight in rural and urban samples, respectively) were greater than the maximum permitted concentration in animal feeding stuffs of plant origin (0.75 WHO-TEQ ng/kg fresh weight, equivalent to 0.85 WHO-TEQ ng/kg dry weight). Notably, differences in dioxin concentrations between rural and urban soils (Table 13) were not reflected by differences in herbage concentration from rural and urban areas (Table 14), highlighting the importance of atmospheric deposition (rather than soil concentrations) to herbage contamination.

Page 33

Table 14. Concentrations of some OCs in crop plants and herbage from England and Wales.

Compound England Wales Crop type Source Dioxins: PCDD/F (WHO-TEQ ng/kg dw) 1.02 0.67 Rural herbage (median) UKSHS (Creaser et al., 2007a) PCDD/F (WHO-TEQ ng/kg dw) 0.90 0.61 Urban herbage (median) UKSHS (Creaser et al., 2007a) Dioxins and dioxin like PCBs (WHO-TEQ ng/kg fw) 0.01 - Vegetables FSA (2007)

PCBs: ΣPCB (26 congeners) (μg/kg dw) 0.89 1.09 Rural herbage (median) UKSHS (Creaser et al., 2007b) ΣPCB (26 congeners) (μg/kg dw 1.55 0.74 Urban herbage (median) UKSHS (Creaser et al., 2007b)

PAHs: ΣPAH (22 PAHs) (μg/kg dw) 101.8 122.3 Rural herbage (median) UKSHS (Creaser et al., 2007c) ΣPAH (22 PAHs) (μg/kg dw) 305 180 Urban herbage (median) UKSHS (Creaser et al., 2007c) BaP (μg/kg dw) 2.62 2.45 Rural herbage (median) UKSHS (Creaser et al., 2007c) BaP (μg/kg dw 20.4 6.18 Urban herbage (median) UKSHS (Creaser et al., 2007c) BaP (μg/kg fw) 0.32 - Cereals FSA (2002) BaP (μg/kg fw) 0.06 - Root vegetables FSA (2002) BaP (μg/kg fw) 0.09 - Other vegetables FSA (2002)

Chlorinated paraffins: Short chain CP (μg/kg fw) 2.0 - Fruit FSA (2009a) Medium chain CP (μg/kg fw) 1.0 - Fruit FSA (2009a)

Perfluoroalkyl substances: PFOS (μg/kg fw) <1 - Vegetables FSA (2009c) PFOA (μg/kg fw) <1 - Vegetables FSA (2009c)

Page 34

3.7 Models

Modelling OC uptake by plants presents various challenges, as there are many different compounds, with greatly varying physico-chemical properties and different patterns of release from the soil. Many OCs are strongly bound to the soil, limiting uptake into plants and our ability to evaluate the influence of soil chemistry on plant uptake. Nevertheless, models to predict plant uptake have been developed ranging from simple screening tools to more complex models that consider plant and soils physical, chemical, and biological processes in a mechanistic manner. Detailed model descriptions and studies of model performance have been published elsewhere (e.g. Trapp and McFarlane, 1994; Fryer and Collins, 2003) and are not repeated in this review.

3.7.1 Screening tools

Screening tools are a simple, widely used methodology for rapidly assessing the potential for OC uptake by plants based on the physico-chemical properties of the chemical. For example, in the US, Ryan et al. (1988) used basic OC properties (half- life, Kow and Henry's Law constant) to screen for potential plant uptake. In the UK, Wild and Jones (1992) used this approach to assess the potential of OCs in sewage sludge to transfer to plants (and livestock). Based on previously published research that established correlations between OC BCFs and Kow, they concluded that lipophilic compounds with high log Kow values (e.g. PAHs, PCBs, PCDD/Fs) were most likely to be adsorbed by the soil and/or plant root. Chemicals with lower Kow values were likely to be translocated within the plant and may reach above ground portions of the plant. Similarly, Duarte-Davidson and Jones (1996) used a screening model to assess the likelihood of OCs accumulating in the food-chain following the application of sewage sludge onto arable and pasture land. Over 300 OCs or groups of OCs were screened, again using the log Kow value to assess the potential for root uptake and translocation within plants as follows:

• High potential for uptake/translocation: 1 < log Kow < 2.5 • Moderate potential for uptake/translocation: 2.5 < log Kow < 3.0 or 0.5 < log Kow < 1.0 • Low potential for uptake/translocation: log Kow < 1.0 or log Kow > 3.0

The authors stated that plant uptake was influenced by soil type, so that plant availability would generally be highest in sandy and low organic matter soils. Thus the simple screening approach probably reflects the ‘worst case’ potential for plant uptake. Although some OCs were predicted to translocate within plants, experimental studies have so far shown little evidence of this actually occurring, especially for volatile/semi-volatile OCs. Based on the above information, a priority list was drawn up of compounds most likely to be taken up by plant roots (although not translocated within the plant) which comprised dioxins, PCBs, heavy chlorobenzenes (including HCB) and heavy chlorophenols (including PCP).

Whilst the screening tool approach is useful in assessing the relative propensity for OC uptake by plants, it does not provide information on predicted concentrations or persistence in different plant parts (and hence on human exposure via food). Paterson et al. (1990) considered there was a need for general plant models which would take into account:

• other OC properties (in addition to Kow); • differences between plant species; Page 35

• uptake into different plant compartments including root, stem, foliage and seeds and/or fruit; and • soil properties e.g. organic matter content, pH.

3.7.2 Compartment models

A ‘simple’ model representing all above ground vegetation as one compartment was developed by Trapp and Matthies (1995). This considered uptake from the soil, atmospheric deposition, volatilization from leaves, transformations and degradation, plant growth and various physico-chemical data on each OC. The model was tested for 2,3,7,8-TCDD (a dioxin) and showed good agreement with experimental data.

There are several models in the literature that divide plants into multiple compartments representing the roots, stem and foliage, and fruit, which have been developed to quantify concentrations in different plant parts for the purpose of exposure assessment. For example, Paterson and Mackay (1989) developed a three-compartment (root, stem and foliage) mass balance model to quantify OC uptake from the soil and atmosphere via:

• diffusion and bulk flow between soil and root; • transport within the plant between root, stem and foliage; • exchange between foliage and air and between soil and air.

The objective was to provide a method by which OC concentrations in various plant tissues could be estimated from information on OC properties, concentrations in soil and air, and plant physiology. The model was applied to uptake of the Bromacil by soybean from hydroponic solution and gave results which compared favourably with experimental data. Application to three other chemicals (2,4-D, dichlorobenzonitrile and hexachlorobiphenyl) showed that OCs present in soil may reach foliage through evaporation and subsequent foliar absorption and transpiration, with the proportions determined by the OC's Henry's law constant and Kow (Paterson and Mackay, 1994).

3.7.3 Fugacity models

Fugacity models are distribution based models incorporating all environmental compartments and are based on steady-state fluxes of contaminants across compartment interfaces. A fugacity model for OCs was proposed by Mackay and Paterson (1981) to calculate the relative amount of an OC that would ultimately partition into each environmental compartment (i.e. air, soil, water and plants).

In its standard form, a fugacity model is basically a screening tool, requiring data on only a few OC physico-chemical properties (e.g. molecular weight, water solubility, vapour pressure and Kow). If the model is applied to the real environment, additional input data include, for example, the volumes of various environmental compartments (e.g. soil volume), temperature, soil characteristics (e.g. % organic carbon) and the quantity of OC introduced. In this form the model is more site-specific and can be used to estimate environmental concentrations. At the next level of complexity, environmental persistence can be evaluated but only where transformation rates of various degradation processes in different environmental compartments (e.g. photolysis, hydrolysis, biodegradation, etc.) are known.

Page 36

Fugacity modelling (Mackay, 2001) can be used to estimate the partitioning of OCs between the various phases of the soil matrix and hence their likelihood of being taken up into plants. The SOIL model (Mackay, 2001) consists of four environmental compartments (i.e. air, water, soil mineral matter and soil organic matter). Densities, volumes, areas and depths of soils are user specified, enabling total volumes and masses to be calculated from which individual fugacities and environmental concentrations can be derived. Indeed, this approach has been used to assess the propensity for OCs in compost applied to soils to enter different soil phases and hence their likelihood of being taken up into plants (WRAP, 2009).

3.7.4 Partition-limited models

Chiou et al. (2001) developed a partition-limited model for the plant uptake of OCs from soil and water. In this model, the maximum (equilibrium) concentration of an OC in any location in the plant is determined by the partition equilibrium with its concentration in soil interstitial water, which in turn is determined by the concentration in soil organic matter. Further, the extent of partition equilibrium (as measured by the ratio of the contaminant concentrations in plant water and soil interstitial water), depends on the transport rate of the OC in soil water into the plant and the volume of soil water solution that is required for the plant contaminant level to reach equilibrium with the external soil-water phase. The model indicated that for plants with high water contents, the plant-water phase acted as the major reservoir for highly water-soluble OCs. In contrast, plant lipids (even in small amounts) were usually the major reservoir for highly water-insoluble OCs.

Zhu and Gao (2004) evaluated the performance of the Chiou et al. (2001) partition- limited model on phenanthrene uptake by a variety of plant species in a greenhouse study. The model predictions of root or shoot concentrations for the tested plant species were all within an order of magnitude of the measured values. Similarly, Gao et al. (2005) used the Chiou et al. (2001) partition-limited model to predict plant uptake of phenanthrene and pyrene from soil and water. Again, the overall model predictions were within an order of magnitude of the measured values. However, in both studies, modelled root concentrations were generally more accurate than modelled shoot concentrations. Very similar results were reported by Gao and Ling (2006).

3.7.5. Summary

There have been worthwhile advances in our understanding of the processes of plant OC uptake in recent years. Collins et al. (2006) concluded that there was sufficient information on the root uptake (and aerial deposition) of OCs to plants to allow for understanding (and hence modelling) of these potential transfer pathways. However, these authors also stated that where relationships have been proposed for use in predictive models, they “have not been subsequently validated by a range of independent experiments as is required for a truly robust model”. Moreover, for some key factors there was still little information available e.g. plant lipid content and plant metabolism.

3.8 Soil ingestion by livestock

The direct ingestion of soil by grazing livestock is considered to be the principal route of OC accumulation in the food chain (Smith, 1996). Wild and Jones (1992) considered that the potential transfer to animal tissues via soil ingestion was dependent on OC persistence in soil and on the bioconcentration potential in Page 37 terrestrial ecosystems (i.e. the lipophilic nature of some ingested OCs – measured by a high Kow - may result in their accumulation in animal tissues, milk and dairy products). These authors defined the following risk categories for the importance of soil ingestion of OCs:

• High potential for soil ingestion: Soil half-life > 21 days and log Kow > 4.5 • Moderate potential for soil ingestion: Soil half-life > 21 days and log Kow > 3.0 • Low potential for soil ingestion: Any other properties.

Dioxins, lighter PCB congeners, higher chlorinated chlorobenzenes and PAHs (except for naphthalene, acenaphthene and fluorene) are all persistent in soil and lipophilic, although PAHs do not appear to accumulate in animal tissue (Fries, 1996). Whilst PCBs, dioxins and CBs are not metabolised by animals and can accumulate in tissue fat and milk, the risks from human consumption of animal products has been assessed to be extremely low (Nicholson et al., 2008). Whilst some concern has been expressed over dioxins and dioxin-like PCB levels in free range eggs, the COT concluded that the data collected by the FSA could not be used to estimate the risk to health of consumers, although it might be used as an indicator of environmental contamination (COT, 2000).

Median dioxin concentrations in herbage in England (0.9 WHO-TEQ ng/kg dw; Table 14), which is likely to be consumed in cattle diets, were greater than the maximum permitted concentration in animal feeding stuffs of plant origin (0.85 ng WHO-TEQ/kg dw; derived from Table 7), although the major route of contamination was likely to be atmospheric deposition rather than uptake from the soil. Importantly, recent data on dioxin (and dioxin like PCB) levels in food (FSA, 2007) showed that mean concentrations in milk and dairy products ranged from 0.42 – 1.30 ng WHO-TEQ/kg fat, which is well below the maximum permitted level of 6 ng WHO-TEQ/kg fat, specified in EC legislation (Table 6), although as this was not an exhaustive survey, it cannot be used to draw conclusions for the UK as a whole.

4. Conclusions

Among the toxic metals/metalloids, there are concentration limits in food (or animal feeding stuff) limits in place for Cd, F, Pb, Hg and Sn, or being considered for As in food. These were used here as an upper limit to assess the significance of concentrations in food and feed, derived from soil. For other metals/metalloids there are no concentration limits in food or feed. Cadmium is the metal that was found to be most likely to be present in soils and taken up by food crops in amounts that are significant in terms of its food limit than other metals/metalloids. Models describing the transfer of Cd from soil to plants and animals have been developed, but no models exist for As and Pb uptake into crops from soils.

Applying models for Cd uptake to the soils in England and Wales shows compliance with the maximum limit concentrations for cereals, potato and herbage in most soils, whereas vegetables may be more problematic if they are grown in soils containing elevated concentrations of Cd in combination with low pH. This is backed up by surveys of cereal grains and herbage show that a very small percentage of samples exceed the current limits for Cd. However, there are no national surveys to back up model predictions for vegetables. Grain surveys also show that few samples exceed the limit for Pb.

If the limits for Cd and Pb in food or feed are lowered in future, this will affect the likely crops and areas that may exceed the limits. This situation needs to be kept Page 38 under review. The potential for uptake and toxicity of nanoparticles, including those that are likely to be deposited on soils (Mueller and Novak, 2008; e.g. those containing Ag, Ce, Ti and Zn) is uncertain at present and the subject of ongoing research.

There are no legislative limit values for OC concentrations in UK soils. Of the OCs for which there are dietary or food limits, many have been detected in UK soils, although levels are generally low in (rural) agricultural soils. Moreover, most OCs are strongly sorbed by and persistent in soil, although there was some evidence that PCP and NP could be leached from or biodegraded (NP only) in soil.

Plant uptake studies indicate that OC transfer from the soil into the above-ground plant parts is likely to be very limited, and that direct deposition from the atmosphere and contamination with dust/soil particles are the major routes of entry. There was some evidence of enrichment of liphophilic OCs (e.g. PCBs) in some root crops (e.g. carrots), but transfers were inefficient and generally confined to the skin, which is normally removed before consumption.

Models of plant OC uptake range from simple screening tools (based on a few OC physicochemical properties such as the Kow) to more complex compartment, fugacity or partition-limited models. Model predictions are confirmed by the low degree of plant OC uptake from soil measured in experimental studies. More complex models are required to predict plant OC concentrations, but these also require much more detailed information about the physicochemical properties of the OC, the plant itself and prevailing soil and environmental conditions.

Finally, it is not clear whether the concentrations of contaminants of interest are increasing or decreasing in soils over time, due to the limited effort in soil monitoring. This is likely to be affected by local conditions including inputs and soil conditions and would require substantial effort to characterise and understand these effects and to accurately detect any changes.

Page 39

References

Adams, M.L., Zhao, F.J., McGrath, S.P., Nicholson, F.A., Chambers, B.J. 2004 Predicting cadmium concentrations in wheat and barley grain using soil properties. J. Environ. Qual. 33, 532-41.

Agency for Toxic Substances and Disease Registry (ATSDR). 1998. Toxicological profile for Chlorinated Dibenzo-p-dioxins (CDDs). Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service.

Agency for Toxic Substances and Disease Registry (ATSDR). 2000. Toxicological profile for Polychlorinated Biphenyls (PCBs). Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service.

Alcock, R.E., Bacon, J., Bardget, R.D., Beck, A.J., Haygarth, P.M. Lee, R.G.M.,. Parker, C.A., and Jones, K.C. 1996. Persistence and fate of polychlroinated biphenyls (PCBs) in sewage sludge amended agricultural soils Environ. Pollut.,. 93, 83-92.

Alcock, R.E., Sweetman, A. and Jones, K.C. 1999. Assessment of organic contaminant fate in waste water treatment plants. I: Selected compounds and physicochemical properties Chemosphere 38, 2247-2262.

Alcock, R.E., Sweetman, A.J., Juan, C.Y. and Jones, K.C. 2000. A generic model of human lifetime exposure to persistent organic contaminants: development and application to PCB-101. Environ. Pollut., 110,. 253-265.

Alexander, P.D., Alloway, B.J., Dourado, A.M. 2006 Genotypic variations in the accumulation of Cd, Cu, Pb and Zn exhibited by six commonly grown vegetables. Environ. Pollut. 144, 736-45.

Ayris, S. and Harrad, S. 1999. The fate and persistence of polychlorinated biphenyls in soil, J. Environ. Monit. 1, 395–401.

Basta NT, Ryan JA, Chaney RL 2005 Trace element chemistry in residual-treated soil: Key concepts and metal bioavailability. Journal of Environmental Quality 34, 49- 63.

Bellin, C. A. and G.A. O'Connor, 1990. Plant uptake of pentachlorophenol from sludge amended soils. J. Environ. Qual., 19, 598-602.

Beyer, A., Mackay, D., Matthies, M., Wania, F. and Webster, E., 2000. Assessing long-range transport potential of persistent organic . Environ. Sci. Technol. 34, 699–703.

Black H.I.J., Garnett J.S., Ainsworth G., Coward P. A., Creamer R., Ellwood S., Horne J., Hornung M., Kennedy V. H., Monson F., Raine L., Osborn D., Parekh N. R., Parrington J., Poskitt J. M., Potter E., Reeves N. Rowland A. P., Self P., Turner S., Watkins J., Woods C.and Wright J. 2002. MASQ: Monitoring and Assessing Soil Quality in Great Britain. Countryside Survey Module 6: Soils and Pollution. R&D Technical Report E1-063/TR. http://nora.nerc.ac.uk/4297/2/SE1-063-TR-e-p.pdf

Page 40

Brus, D.J., de Gruijter, J.J., Walvoort, D.J.J., de Vries, F., Bronswijk, J.J.B., Romkens, P., de Vries, W. 2002 Mapping the probability of exceeding critical thresholds for cadmium concentrations in soils in the Netherlands. J. Environ. Qual. 31, 1875-84.

Brus, D.J., Jansen, M.J. W. 2004 Uncertainty and sensitivity analysis of spatial predictions of heavy metals in wheat. J. Environ. Qual. 33, 882-90.

Casterline, J. L. Jr, Barnett, N. M., Ku, Y. 1985. Uptake, translocation, and transformation of pentachlorophenol in soybean and spinach plants. Environ Res. 37, 101-18.

Chang A. C., Pan, G., Page, A.L. and Takashi Asano, T. 2002. Developing Human Health-related Chemical Guidelines for Reclaimed Water and Sewage Sludge Applications in Agriculture. Prepared for World Health Organization. http://www.who.int/water_sanitation_health/wastewater/gwwuchemicals.pdf (accessed 17/03/10)

Chaudri, A.M., Zhao, F.J., McGrath, S.P., Crosland, A.R. 1995 The Cadmium content of British wheat grain. J. Environ. Qual. 24, 850-5.

Chaudri, A.M., McGrath, S.P., Gibbs, P.A., Chambers, B.J., Carlton-Smith, C.H., Godley, A., Bacon, J.R., Campbell, C.D. and Aitken, M.N. 2007. Cadmium availability to wheat grain in soils treated with sewage sludge or metal salts. Chemosphere 66, 1415–1423.

Chiou, C.T., Sheng, G. and Manes, M., 2001. A partition-limited model for the plant uptake of organic contaminants from soil and water. Environ. Sci. Technol. 35, 1437– 1444.

Collins, C., Fryer, M., Grosso, A., 2006. Plant uptake of non-ionic organic chemicals. Environ. Sci. Technol. 40, 45-52.

Combs, G. F. 2001 Selenium in global food systems. British Journal of Nutrition 85, 517-47.

Commission of the European Communities 2001 Commission Regulation (EC) No 466/2001 of 8 March 2001 setting maximum levels for certain contaminants in foodstuffs. Official Journal L 077 , 16/03/2001 P. 0001 – 0013, Brussels.

COT 2000. Statement on dioxins and dioxin-like PCBs in free-range eggs (June 2000). http://cot.food.gov.uk/cotstatements/cotstatementsyrs/cotstatements2000/dioxins_eg gs

COT 2001. Statement on the Tolerable Daily Intake for Dioxins and Dioxinlike Polychlorinated Biphenyls. Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment. http://www.food.gov.uk/multimedia/pdfs/cot-diox- full.pdf

Cousins, I.T. and Jones, K. C. 1998. Air–soil exchange of semivolatile organic compounds in the UK. Environ. Pollut. 102, 105–118.

Page 41

Cousins, I.T.,Gevao, B. and Jones, K.C.1999 Measuring and modelling the vertical distribution of semi-volatile organic compounds in soil. I: PCB and PAH soil core data. Chemosphere 39: 2507-2518.

Creaser C. S., Wood M. D., Alcock, R., Copplestone, D. Crook, P. J. 2007a. UK Soil and Herbage Survey. Environmental Concentrations of polychlorinated dibenzo-p- dioxins and polychlorinated dibenzofurans in UK Soil and Herbage. UKSHS Report Number 10. Environment Agency, Bristol.

Creaser C. S., Wood M. D., Alcock, R., Copplestone, D. Crook, P. J. 2007b. UK Soil and Herbage Survey. Environmental Concentrations of Polychlorinated Biphenyls (PCBs) in UK Soil and Herbage. UKSHS Report Number 8. Environment Agency, Bristol.

Creaser C. S., Wood M. D., Alcock, R., Copplestone, D. Crook, P. J. 2007c. UK Soil and Herbage Survey. Environmental Concentrations of Polyaromatic hydrocarbons (PAHs) in UK Soil and Herbage. UKSHS Report Number 9. Environment Agency, Bristol

Dalla Valle, M. Marcomini, A., Sweetman, A. J. and Jones and K. C. 2005. Temporal trends in the sources of PCDD/Fs to and around the Venice Lagoon, Environ Int 31, 1040–1046.

Dalenberg, J. W., Vandriel, W. 1990 Contribution of atmospheric deposition to heavy metal concentrations in field crops. Neth. J. Agric. Sci. 38, 369-79.

Defra and EA. 2002a Soil guideline values for cadmium contamination. Bristol: Environment Agency, 1-15.

Defra and EA. 2002b Soil guideline values for lead contamination. Bristol: Environment Agency, 1-15.

Defra 2009a Potential health effects of contaminants in soil. Final report project SP1002. www.defra.gov.uk

Defra 2009b Safeguarding our Soils. A Strategy for England. www.defra.gov.uk

DoE 1996. Code of Practice for Agricultural Use of Sewage Sludge. London, Department of the Environment.

Doick, K. J., Klingelmann, E., Burauel, P., Jones, K. C. and Semple, K. T. 2005. Long-term fate of polychlorinated biphenyls and polycyclic aromatic hydrocarbons in an agricultural soil, Environ. Sci. Technol. 39, 3663–3670. di Domenico, A., Silano, V., Viviano, G. and Zapponi, G. 1980. Accidental release of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) at Sèveso, Italy. II. TCDD distribution in the soil surface layer. Ecotoxicol. Environ. Saf. 4, 298–320.

Donaldson, S.G. and Miller, G.C. 1997. Transport and photolysis of PCP in soils subject to evaporating water. J. Env. Quality, 26, 402–409.

Page 42

Duarte-Davidson, R. and Jones, K.C. 1996. Screening the environmental fate of organic contaminants in sewage sludge applied to agricultural soils: II. The potential for transfers to plants and grazing animals. Sci. Total Environ. 185, 59-70

EC 2000. Working Document on Sludge, Third Draft (ENV.E.3/LM April 2000)

EC. 2001a. Opinion of the Scientific Committee on Food on the risk assessment of dioxins and dioxin-like PCBs in food. Update based on new scientific information available since the adoption of the SCF opinion of 22nd November 2000 (adopted on 30 May 2001). http://ec.europa.eu/food/fs/sc/scf/out90_en.pdf

EC 2001b. Council Regulation (EC) No 2375/2001 of 29 November 2001 amending Commission Regulation (EC) No 466/2001 setting maximum levels for certain contaminants in foodstuffs. http://eur- lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2001:321:0001:0005:EN:PDF (accessed 30/3/11)

EC. 2001c. Commission Regulation (EC) No 466/2001 of 8 March 2001 setting maximum levels for certain contaminants in foodstuffs. http://www.caobisco.com/doc_uploads/legislation/466-2001EN.pdf (accessed 301/3/11)

EC. 2006. Commission Regulation 1881/2006 setting maximum levels for certain contaminants in foodstuffs. Official Journal of the European Union. L 364/5.

EFSA (European Food Safety Authority). 2005. Opinion of the Scientific Panel on contaminants in the food chain [CONTAM] related to the presence of non dioxin-like polychlorinated biphenyls (PCB) in feed and food. http://www.efsa.europa.eu/en/efsajournal/pub/284.htm

EFSA (European Food Safety Authority). 2008 Scientific Opinion of the Panel on Contaminants in the Food Chain on Polycyclic Aromatic Hydrocarbons in Food. EFSA Journal 724, 1-114

EFSA (European Food Safety Authority). 2009a Scientific opinion on cadmium in food. EFSA Journal 980, 1-39.

EFSA (European Food Safety Authority). 2009b Scientific opinion on arsenic in food. EFSA Journal 7: 1351, 1-199.

Environment Agency, 2009a. Using Soil Guideline Values. Science report: SC050021/SGV introduction http://www.grdp.org/static/documents/Research/SCHO0309BPQM-e-e.pdf (accessed 23/03/10)

Environment Agency, 2009b. Updated technical background to the CLEA model. Science Report SC050021/SR3. Bristol: Environment Agency.

Environment Agency. 2009c Soil guideline values for inorganic arsenic in soil. Bristol: Environment Agency, 1-11.

EMEP. 2009. Detailed reports for each EMEP country for 2009: UK. http://www.msceast.org/countries/United_Kingdom/index.html (accessed 31/3/10)

Page 43

Engwall, M. and Hjelm, K. 2000. Uptake of dioxin-like compounds from sewage sludge into various plant species – assessment of levels using a sensitive bioassay, Chemosphere 40, 1189–1195

Eurochlor. 2010. Chlorinated paraffins – Environmental Impact http://www.eurochlor.org/environmentalimpact

FAO/UNEP. 1996. Joint FAO/UNEP Programme For The Operation Of Prior Informed Consent. Decision Guidance Documents. Pentachlorophenol and its salts and esters. http://www.pic.int/en/DGDs/PentachlophEN.doc (accessed 01/03/10)

FAO/WHO 2009. Code Alimentarius. Pesticide Residues in Food. http://www.codexalimentarius.net/mrls/pestdes/jsp/pest_q-e (accessed 17/03/10).

Fries, G. F. 1996. Ingestion of sludge applied organic chemicals by animals. Sci. Tot. Environ. 185, 93-108.

Fryer, M. E. and Collins, C. D., 2003. Model intercomparison for the uptake of organic chemicals by plants. Environ. Sci. Technol. 37, 1617–1624.

FSA. 2002. PAHs in total diet. http://www.food.gov.uk/multimedia/pdfs/31pah.pdf (accessed 17/03/10)

FSA. 2007. Dioxins And Dioxin-Like PCBs In Foods – EU Monitoring 2006. FSIS04/07. http://www.food.gov.uk/multimedia/pdfs/fsis0407.pdf (accessed 17/03/10).

FSA. 2009a. Food Contaminants – Update Bulletin November 2009.

FSA. 2009b Measurement of the concentrations of metals and other elements from the 2006 UK total diet study. Food Survey Information Sheet 01/09 http://www.food.gov.uk/science/surveillance/fsisbranch2009/survey0109. London: Food Standards Agency.

FSA. 2009c. Survey Of Fluorinated Chemicals In Food. Food Survey Information Sheet Number 05/09 October 2009. http://fluoridealert.org/re/pfoa- pfos.uk.survey.2009.pdf (accessed 01/03/10).

FSA 2010. Food Contact Materials, Contaminants and Irradiation – Update Bulletin August 2010. http://www.food.gov.uk/multimedia/pdfs/contaminantsbulletinaug10.pdf

Gao, Y., Zhu, L.,and Ling, W. 2005. Application of the partition-limited model for plant uptake of organic chemicals from soil and water. Sci. Tot. Env. 336, 171-182.

Gao, Y. and Ling, W. 2006. Comparison for plant uptake of phenanthrene and pyrene from soil and water. Biol. Fertil. Soils 42, 387-394.

Ghanem, A., Dubroca, J., Chaplain, V. and Mougin, C. 2006. Fate of and nonylphenol in soil–plant–water systems amended with contaminated sewage sludge Environ. Chem. Letters, 4, 63-67.

Gibbs, P.A., Chambers, B.J., Chaudri, A.M., McGrath, S.P., Carlton-Smith, C.H., Bacon, J.R., Campbell, C.D., and Aitken, M.N. 2006. Initial results from a long-term, Page 44 multi-site field study of the effects on soil fertility and microbial activity of sludge cakes containing heavy metals. Soil Use and Management 22, 11-21.

Gibson, R., Wang, M.-J., Padgett, E. and Beck, A. J. 2005. Analysis of 4- nonylphenols, phthalates, and polychlorinated biphenyls in soils and biosolids. Chemosphere 61, 1336-1344

Giesy, J. P. and Kannan, K. 2002. Perfluorochemical surfactants in the environment, Environ. Sci. Technol. 36,147A–151A.

Gouin, T., Mackay, D., Jones, K. C., Harner, T. and Meijer, S. N. 2004. Evidence for the ‘‘grasshopper’’ effect and fractionation during long-range atmospheric transport of organic contaminants. Environ. Pollut. 128, 139–148

Haluska, L., Barancíková, G., Baláz, S., Dercová, K., Vrana, B., Paz-Weisshaar, M., Furciova, E. and Bielek, P. 1995. Degradation of PCB in different soils by inoculated Alcaligenes xylosoxidans. Sci. Tot. Env. 175, 275-285.

Harms, H.H., 1996. Bioaccumulation and metabolic fate of sewage sludge derived organic xenobiotics in plants. Sci. Total Environ. 185, 83-92.

Hillman, J.P.,Hill, J.,Morgan, J.E.,Wilkinson, J.M. 2003. Recycling of sewage sludge to grassland: a review of the legislation to control of the localization and accumulation of potential toxic metals in grazing systems. Grass and Forage Science 58, 101-111.

Hough, R.L., Breward, N., Young, S.D., Crout, N.M.J., Tye, A.M., Moir, A.M. and Thornton, I. 2004. Assessing potential risk of heavy metal exposure from consumption of home-produced vegetables by urban populations. Environmental Health Perspectives 112,215-221.

Hülster, A., Müller, J.F. and Marschner, H. 1994. Soil-plant transfer of polychlorinated dibenzo-p-dioxins and dibenzofurans to vegetables of the cucumber family (Cucurbitaceae). Environ. Sci. Technol. 28, 1110-1115.

Imperial College Consultants 2001. Pollutants in Urban Waste Water and Sewage Sludge. Final Report to DG Environment. ICON, London.

Jones, K. C. and Stevens, J. (2002). Organic Contaminants in Sewage Sludge Applied to Agricultural Land. Final report to UKWIR for project 02/SL/04/2.

Lead, W. A., Steinnes, E., Bacon, J. R. and Jones, K. C. 1997. Polychlorinated biphenyls (PCBs) in UK and Norwegian soils: spatial and temporal trends. Sci. Total Environ., 193,. 229-236.

Li, F., Zhang, C., Qu, Y., Chen, J., Chen, L., Liu, Y., Zhou, Q. 2010. Quantitative characterization of short- and long-chain perfluorinated acids in solid matrices in Shanghai, China. Sci. Tot. Environ. 408, 617–623.

Li HF, Gray C, Micό C, Zhao FJ, and McGrath SP 2008 Phytotoxicity and bioavailability of cobalt to plants in a range of soils. Chemosphere 75, 979-986.

Page 45

Kipopoulou, A. M., Manoli, E. and Samara, C., 1999. Bioconcentration of polycyclic aromatic hydrocarbons in vegetables grown in an industrial area. Environ. Pollut. 106, 369-380.

Kuboi, T., Noguchi, A., Yazaki, J. 1986 Family dependent cadmium accumulation characteristics in higher plants. Plant Soil 92, 405-15.

Mackay, D. 2001. Multimedia Environmental Models: The Fugacity Approach 2nd edition. Lewis Publishers, London.

Mackay, D. and Paterson, S. 1981. Calculating fugacity. Environ. Sci. Technol. 15, 1006-1014.

Mackay, D., Shiu, W. Y. and Ma, K. C., 1992. Illustrated Handbook of Physical- Chemical Properties and Environmental Fate of Organic Chemicals. Boca Raton, FL: Lewis Publishers.

McGrath, S.P. 1987. Long term studies of metal transfers following applications of sewage sludge. In: Pollutant Transport and Fate in Ecosystems. Eds. PJ Coughtrey, MH Martin and MH.Unsworth. Special Publication No.6 of the British Ecological Society, Oxford, Blackwell Scientific, pp.301-317.

McGrath, S.P. and Loveland, P.J. 1992. The Soil Geochemical Atlas of England and Wales. Glasgow, Blackie. de Meeûs, C., Eduljee, G. H., Hutton, M. 2002 Assessment and management of risks arising from exposure to cadmium in fertilisers. I. Sci. Total Environ. 291, 167-87.

Meharg, A.A., Williams, P.N., Adamako, E., Lawgali, Y.Y., Deacon, C., Villada, A., Cambell, R.C.J., Sun, G., Zhu, Y.G., Feldmann, J., Raab, A., Zhao, F.J., Islam, R., Hossain, S., Yanai, J. 2009. Geographical variation in total and inorganic arsenic content of polished (white) rice. Environmental Science and Technology 43, 1612- 1617.

Meijer, S.N., Ockenden, W.A., Sweetman, A., Breivik, K., Grimalt, J.O. and Jones, K.C., 2003. Global distribution and budget of PCBs and HCB in background surface soils: implications for sources and environmental processes. Environ. Sci. Technol. 37, 667–672.

Mills, T., Arnold, B., Sivakumaran, S., Northcott, G., Vogeler, I., Robinson, B., Norling, C. and Leonil, D. 2006. and long-term site management of soil contaminated with pentachlorophenol (PCP) and heavy metals. J. Environ. Manag. 79, 232-241.

Moir, A. M., Thornton, I. 1989. Lead and cadmium in urban allotment and garden soils and vegetables in the United Kingdom. Environ. Geochem. Health 11, 113-9.

Mueller N.C., and Nowack B. 2008. Exposure modeling of engineered nanoparticles in the environment. Environ. Sci. Technol. 2008, 42 (12), 4447-4453.

Nam, J. J., Gustafsson, O., Kurt-Karakus, P., Breivik, K., Steinnes, E. and Jones, K. C. 2008. Relationships between organic matter, black carbon and persistent organic

Page 46 pollutants in European background soils: Implications for sources and environmental fate. Environ. Pollut. 156, 45-48.

Nicholls, C. R., Allchin, C. R. and Law, R. J., 2001. Levels of short and medium chain length polychlorinated n-alkanes in environmental samples from selected industrial areas in England and Wales, Environ Pollut 114, 415–430.

Nicholson, F.A., Smith, S.R., Alloway, B.J., Carlton-Smith, C., Chambers, B.J. 2003. An inventory of heavy metal inputs to agricultural soils in England and Wales. Science of the Total Environment 311, 205-219.

Nicholson, F.A., Chambers, B.J., Smith, S.R., Riddell-Black, D., Butler, A.P., Rose, S.C. and Lamb, R. 2008. Sources and Impacts of Past, Current and Future Contamination of Soil. Final report for Defra Project SP0547.

Nielsen, E., Østergaard, G., Thorup, I., Ladefoged, O., Jelnes, O. and Jelnes, J. E. 2000. Toxicological evaluation and limit values for nonylphenol, nonylphenol ethoxylates, tricresyl, phosphates and benzoic acid, Environmental Project vol. 512, Danish Environmental Protection Agency, Copenaghen, DK, p. 43.

O’Connor, G. A., Kiehl, D., Eiceman, G. A. and Ryan, J.A. 1990. Plant uptake of sludge-borne PCBs. J. Environ. Qual. 19, 113-118.

OECD 2002. Co-Operation on Existing Chemicals. Hazard Assessment of Perfluorooctane Sulfonate (PFOS) and Its Salts http://www.oecd.org/dataoecd/23/18/2382880.pdf (accessed 17/03/10).

Park, K.S., Sims, R.C., Dupont, R.R., Doucette, W.J., Matthews, J.E., 1990. Fate of PAH compounds in two soil types: influence of volatilization, abiotic loss and biological activity. Environ. Toxicol.Chem. 9, 187-195.

Paterson, S., Mackay, D., Tam, D. and Shiu, W. Y. 1990. Uptake of organic chemicals by plants: A review of processes, correlations and models. Chemosphere, 21, 297-331, 1990.

Paterson, S. and Mackay, D. 1989. Modelling the uptake and distribution of organic chemicals in plants. In: D.T. Allen and I.R. Kaplan (eds), Intermedia Pollutant Transport: Modelling and Field Measurements. Plenum Pub. Co.

Paterson, S. and Mackay, D., 1994. A model of organic chemical uptake by plants from soil and the atmosphere. Environ. Sci. Technol. 28, 2259–2266.

Rayman, M.P. 2002 The argument for increasing selenium intake. Proc. Nutr. Soc. 61, 203-15.

Rhind, S.M., Kyle, C.E. and Owen, J. 2005a. Accumulation of heavy metals in the liver tissue of sheep grazed on sewage sludge-treated pastures., Animal Science, 81, 107-113.

Rhind, S.M., Kyle, C.E., Telfer, G., Duff, E.I. and Smith, A. 2005b. Alkyl phenols and diethylhexyl phthalate in tissues of sheep grazing pastures fertilized with sewage sludge or inorganic fertilizer., Environmental Health Perspectives, 113, 447-453

Page 47

Roberts, P., Roberts, J.P. and Jones, D.L. 2006. Behaviour of the endocrine disrupting chemical nonylphenol in soil: Assessing the risk associated with spreading contaminated waste to land. Soil Biol. Biochem. 38, 1812–1822.

Rooney, C.P., Zhao, F.J., and McGrath, S.P. 2006. Soil factors controlling the expression of copper toxicity to plants in a wide range of European soils. Environmental Toxicology and Chemistry 25, 726-732.

Ross, S.M,, Wood M.D., Copplestone, D., Warriner, M. and Crook, P. 2007 Environmental concentrations of heavy metals in UK soil and herbage. UKSHS Report Number 7. Environment Agency, Bristol.

Ryan, J.A., Bell, R.M., Davidson, J.M. and O’Connor, G.A. 1988. Plant uptake of non- ionic organic chemicals from soil, Chemosphere, 17, 2299-2323.

Scelza, R., Rao, M.A. and Gianfreda, L. 2008. Response of an agricultural soil to pentachlorophenol (PCP) contamination and the addition of compost or dissolved organic matter. Soil Biol. Biochem. 40, 2162-2169.

SI 2005. Statutory Instrument No. 3281. The Feeding Stuffs (England) Regulations 2005. www.opsi.gov.uk/si/si2005/20053281.htm (accessed 27/01/10).

SI 2009. Statutory Instrument No. 1223. The Contaminants in Food (England) Regulations 2009. http://www.opsi.gov.uk/si/si2009/uksi_20091223_en_1 (accessed 16/03/10)

Simonich, S. and Hites, R. 1995. Organic pollutant accumulation in vegetation. Environ. Sci. Technol. 29, 2905-2914.

Sinkkonen, S. and Paasivirta, J. 2000. Degradation half-life times of PCDDs, PCDFs, and PCBs for environmental fate modelling, Chemosphere 40, 943–949.

Smith, S. R. 1996. Agricultural Recycling of Sewage Sludge and the Environment. CAB International.

Soares, A., Guieysse, B., Jefferson, B., Cartmell, E. Lester, J. N. 2008. Nonylphenol in the environment: A critical review on occurrence, fate, toxicity and treatment in wastewaters Environ Int. 34, 1033–1049.

Stahl, T.; Heyn, J., Thiele, H., Hüther, J.; Failing, K.; Georgii, S. 2009. Carryover of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) from soil to plants. Arch. Environ Contamin. Toxicol. 57, 289-298.

Stark, B.A. and Hall, J.E. 1992. Implications of sewage sludge application to pasture on the intake of contaminants by grazing animals. In: Effects of Organic Contaminants in sewage sludge on Soil Fertility. Eds: Hall, J.E., Sauerbeck, D.R., and L’Hermite, P. Comission of the European Communities , Luxemburg. Pp134-157.

Stevens, J. L., Northcott, G. L., Stern, G. A., Tomy, G. T. and Jones K. C. 2003. PAHs, PCBs, PCNs, organochlorine pesticides, synthetic musks and polychlorinate n-alkenes in UK sewage sludge : survey results and implications. Environmental Science and Technology 37, 462-467.

Page 48

Stevens, J. L. and Jones, K. C. 2003. Quantification of PCDD/F concentrations in animal manure and comparison of the effects of the application of cattle manure and sewage sludge to agricultural land on human exposure to PCDD/Fs. Chemosphere 50, 1183-1191.

Tao, S., Cui, Y.H., Xu, F.L., Li, B.G., Cao, J., Liu, W.X., Schmitt, G., Wang, X.J., Shen, W.R., Qing, B.P. and Sun, R. 2004. Polycyclic aromatic hydrocarbons (PAHs) in agricultural soil and vegetables from Tianjin. Sci Total Environ 320, 11–24

Thompson, R.S., Smyth, D.V. and Gillings, E. 2005. Medium-chain chlorinated paraffins (C14-17, 52% chlorinated): Accumulation from soil by the root of carrot (Daucus carota). Brixham Environmental Laboratory, AstraZeneca UK Limited, Study Number BL8221/B.

Thomson, I., Thornton, I., Webb, J.S. 1972 Molybdenum in black shales and incidence of bovine hypocuprosis. J. Sci. Food Agric. 23, 879-891.

Thornton, I. and Abrahams, P. 1983 Soil ingestion - a major pathway of heavy metals into livestock grazing contaminated land. The Science of the Total Environment 28: 287-294.

Tjell, J.C., Hovmand, M.F., Mosbaek, H. 1979 Atmospheric lead pollution of grass grown in a background area in Denmark. Nature 280, 425-6.

Topp, E., Scheunert, I., Attar, A., Korte, F. 1986. Factors affecting the uptake of 14C- labelled organic chemicals by plants from soil. Ecotoxicol. Environ. Saf. 11, 219–228.

Topp, E. and Starratt, A., 2000. Rapid mineralisation of the endocrine disrupting chemical 4-nonylphenol in soil. Environ. Toxicol. Chemistry 19, 313–318.

Trapp, S. and McFarlane, J. C. (Eds) 1994. Plant contamination: modeling and simulation of organic chemical processes. CRC Press.

Trapp, S. and Matthies, M. 1995. Generic one-compartment model for uptake of organic chemicals by foliar vegetation. Environ. Sci. Technol. 29, 2333–2338.

Tremel, A., Masson, P., Garraud, H., Donard, O. F. X., Baize, D., Mench, M. 1997 Thallium in French agrosystems.2. Concentration of thallium in field-grown rape and some other plant species. Environ. Pollut. 97, 161-8.

US EPA; US Environmental Protection Agency 2003. Agency Final Action not to Regulate Dioxins in Land-Applied Sewage Sludge. EPA-822-F-03-007, October 2003. US EPA, Office of Water, Washington DC. (http://www.epa.gov/ost/biosolids/dioxinfs.html)

Walker, C.H. 2001. Organic Pollutants – An Ecotoxicological Perspective. Taylor & Francis, ISBN 0 7484 0962 9.

Wang, M.J. and Jones, K.C., 1994. Uptake of chlorobenzenes by carrots from spiked and sewage-amended soil. Environ. Sci. Technol. 28, 1260–1267.

Page 49

Weeks, C. A., Brown, S. N., Vazquez, I., Thomas, K., Baxter, M., Warriss, P. D., Knowles, T. G. 2007 Multi-element survey of allotment produce and soil in the UK. Food Additives and Contaminants 24, 877-85.

White, P.J., Broadley, M.R. 2009 Biofortification of crops with seven mineral elements often lacking in human diets - iron, zinc, copper, calcium, magnesium, selenium and iodine. New Phytol. 182, 49-84.

WHO, 1990. Evaluation of certain food additives and contaminants. Thirty-fifth report of the JointFAO/WHO Expert Committee on Food Additives. Geneva, World Health Organization, 1990 (Technical Report Series, No. 789).

WHO 1996. Chlorinated paraffins. Geneva, World Health Organization (Environmental Health Criteria No. 181).

WHO 1997. Hexachlorobenzene. Geneva, World Health Organization (Environmental Health Criteria, No. 195).

WHO 1998. WHO revises the tolerable daily intake (TDI) for dioxins. World Health Organization.18th Symposium on Halogenated Environmental Organic Pollutants. Stockholm, Sweden. August 17-21,1998. Organohalogens Compounds. Volume 38, 295-298.

WHO 2000. Air Quality Guidelines, 2nd Edition. Chapter 5.10: PCBs. (http://www.euro.who.int/document/aiq/5_10pcbs.pdf)

WHO 2003. Health Risks of Persistent Organic Pollutants from Long-Range Transboundary . Joint WHO/Convention Task Force On The Health Aspects Of Air Pollution. http://www.euro.who.int/Document/e78963.pdf (accessed 16/03/10)

Wild, S.R. and Jones, K.C. 1992. Organic chemicals entering agricultural soils in sewage sludges: Screening for their potential to transfer to crop plants and livestock. Sci. Tot. Environ. 119, 85–119.

Wild, S.R., Berrow, M.L. and Jones, K.C. 1991. The persistence of polynuclear aromatic hydrocarbons (PAHs) in sewage sludge amended agricultural soils. Environ. Pollut. 72, 141-57.

Wild, S.R., Beck, A.J. and Jones, K.C. 1995. Predicting the fate of non-ionic organic chemicals entering agricultural soils following sewage sludge application. Land Contamin. Reclam. 3, 181-190.

Williams, P.N., Villada, A., Deacon, C., Raab, A., Figuerola, J., Green, A.J., Feldmann, J., Meharg, A.A. 2007 Greatly enhanced arsenic shoot assimilation in rice leads to elevated grain levels compared to wheat and barley. Environ. Sci. Technol 41, 6854-9.

WRAP 2009 Risk Assessment of Application of Composted Wastes to Land. Final report by Macaulay Land Use Research Institute.

Yang, Y., Hunter, W., Tao, S. Gan, J. 2009. Microbial Availability of Different Forms of Phenanthrene in Soils. Environ. Sci. Technol. 43, 1852–1857.

Page 50

Zhao, F.J., Adams, M.L., Dumont, C., McGrath, S.P., Chaudri, A.M., Nicholson, F.A., Chambers, B.J., Sinclair, A.H. 2004 Factors affecting the concentrations of lead in British wheat and barley grain. Environ. Pollut. 131, 461-8.

Zhao, F.J., Stroud, J.L., Eagling, T., Dunham, S.J., McGrath, S.P., Shewry, P.R. 2010a The accumulation, distribution and speciation of arsenic in wheat grain. Environ. Sci. Technol., 44, 5464-5468.

Zhao, F.J., McGrath, S.P., Meharg, A.A. 2010b Arsenic as a food-chain contaminant: mechanisms of plant uptake and metabolism and mitigation strategies. Ann. Rev. Plant Biol. 61, in press, doi:10.1146/annurev-arplant-042809-112152.

Zhu, L. and Gao, Y. 2004. Prediction of phenanthrene uptake by plants with a partition-limited model. Environ. Pollut. 131, 505-508.

Page 51

Annex 1. FAO/WHO Codex Alimentarius Standards - heavy metalsa

Element Commodity Maximum limit mg/kg wet weight

Arsenic (inorganic) Only fats and oils, mineral water, salt - Cadmium Brassica vegetables 0.05 Bulb vegetables 0.05 Fruiting vegetables, cucurbits 0.05 Fruiting vegetables, other than cucurbits 0.05 Leafy vegetables 0.2 Legume vegetables 0.1 Peeled potato 0.1 Pulses 0.1 Root and tuber vegetables 0.1 Stalk and stem vegetables 0.1 Cereal grains, except buckwheat, cañihua 0.1 and quinoa, excluding bran and germ Rice, polished 0.4 Wheat 0.2

Lead Berries and other small fruit 0.2 Citrus fruits 0.1 Pome fruits 0.1 Stone fruits 0.1 Brassica vegetables 0.3 Bulb vegetables 0.1 Fruiting vegetables, cucurbits 0.1 Fruiting vegetables, other than cucurbits 0.1 Leafy vegetables, excluding spinach 0.3 Legume vegetables 0.2 Pulses 0.2 Root and tuber vegetables, peeled potatoes 0.1 Cereal grains, except buckwheat, cañihua 0.2 and quinoa Meat of cattle, pigs, sheep and poultry 0.1 Cattle, edible offal 0.5 Pig, edible offal 0.5 Poultry edible offal 0.5 Milk 0.02

Mercury Only mineral water, salt -

Tin Only canned foods, as Sn comes from - containers and solders, not soil

a General Standard for Contaminants and Toxins in Food and Feed, Codex Standard Volume 1a, 193-1995, Revision 5; 2009. Excluding seafood.

Page 52

Annex 2. EU Food Regulation 1881/2006 - heavy metalsa

Commodity Maximum limit mg/kg wet weight

Arsenic None -

Cadmium Meat of cattle, pigs, sheep and poultry 0.05 Horsemeat 0.2 Liver of bovine animals, sheep, pig, poultry and horse 0.5 Kidney of bovine animals, sheep, pig, poultry and horse 1 Cereals excluding bran, germ, wheat and rice 0.1 Bran, germ, wheat and rice 0.2 Soybeans 0.2 Vegetables & fruit excluding leaf, root and stem vegetables, fresh herbs, fungi, 0.05 pine nuts, and potatoes Leaf vegetables, fresh herbs. Cultivated fungi and celeriac 0.2 Stem vegetables, root vegetables and peeled potatoes 0.1

Lead Milk 0.02 Meat of cattle, pigs, sheep and poultry 0.1 Offal of bovine animals, sheep, pigs and poultry 0.5 Cereals, legumes and Pulses 0.2 Vegetables and peeled potatoes, excluding brassicas, leaf vegetables, 0.1 herbs and fungi Brassica vegetables, leaf vegetables and cultivated fungi 0.3 Fruit excluding berries and small fruit 0.1 Berries and small fruit 0.2

Mercury Only seafood limits -

Tin Only canned food -

Above Regulation enforced as Statutory Instrument No. 1223, The Contaminants in Food (England) Regulations 2009. a Excluding seafood.

Page 53

Annex 3. EU Feed Regulation - heavy metalsa

Maximum Maximum mg/kg at 88% DM mg/kg DM Arsenic Feed 2 2.3

Cadmium Feed 1 1.1

Lead Green fodder 40 45

Fluorine Feed 150 170

Mercuryb Feed 0.1 0.11

aDirectives 2002/32/EC and amendment 2003/100/EC. bHg added in No. 3281 Agriculture The Feeding Stuffs (England) Regulations 2005.

Page 54