Cd, Pb and Zn Oral Bioaccessibility of Urban Soils Contaminated in the Past by Atmospheric Emissions from Two Lead and Zinc Smelters
- First Online:
- Cite this article as:
- Roussel, H., Waterlot, C., Pelfrêne, A. et al. Arch Environ Contam Toxicol (2010) 58: 945. doi:10.1007/s00244-009-9425-5
Ingestion of dust or soil particles could pose a potential health risk due to long-term metal trace element (MTE) exposure. Twenty-seven urban topsoil samples (kitchen garden and lawn) were collected and analyzed for Cd, Pb and Zn using the unified Bioaccessibility Research Group of Europe (BARGE) method (UBM) test to estimate the human bioaccessibility of these elements. The quantities of Cd, Pb and Zn extracted from soils indicated, on average, 68, 62 and 47% bioaccessibility, respectively, in the gastric phase and 31, 32 and 23% bioaccessibility, respectively, in the gastro-intestinal phase. Significant positive correlations were observed between concentrations extracted with UBM and total MTE contents. Stepwise multiple regression analysis showed that human bioaccessibility was also affected by some physico-chemical soil parameters (i.e. total nitrogen, carbonates, clay contents and pH). The unified test presents some valuable data for risk assessment. Indeed, the incorporation of oral bioaccessible concentrations into risk estimations could give more realistic information for health risk assessment.
Soil metallic pollution has dramatically increased with anthropogenic actions. In Northern France, metal trace element (MTE) soil contamination due to the past activities of two lead and zinc smelters may present a significant health risk for the surrounding population. Also, governmental and regional institutions are devoting increased attention to soil pollution problems, which are closely related to soil use. Indeed, urban soils present peculiar characteristics compared with agricultural soils, resulting from high anthropogenic pressure, which complicates investigations (de Kimpe and Morel 2000). In several studies, the contamination levels of urban soils have often been shown to be higher than those of their agricultural counterparts (Paterson et al. 1996; Chen et al. 1997; Pruvot et al. 2006; Douay et al. 2008). The pedological evolution of these soils is dominated by their history as well as their past and current uses. Besides, the specificities of the physico-chemical parameters of urban soils could lead to metal behaviours different from those observed in agricultural soils. The physico-chemical parameters of urban soils are generally known to be strongly affected by anthropogenic actions, the outcomes of which include compaction of deeper layers, low biodiversity, presence of fragments of various materials and often high concentrations of pollutants. Indeed, these urban soils are over the years the final repository of diverse elements and potentially toxic wastes (Chirenje et al. 2002; Norra and Stüben 2003; Peltola and Astrom 2003; Alloway 2004). It has been reported that in some gardens slag was used as a herbicide, pieces of bricks were added to embankments and ashes originating from domestic coal combustion were spread onto the soil (Douay et al. 2007, 2008).
Environmental research carried out around two lead and zinc smelters in Northern France (Metaleurop Nord at Noyelles-Godault and Umicore at Auby) has shown high Cd, Pb and Zn contamination in agricultural and urban topsoils (Sterckeman et al. 2000, 2002; Douay et al. 2002, 2005, 2008). Metaleurop Nord, which was the largest producer of primary lead in Europe, closed in 2003 after more than a century of pyrometallurgic processing that generated large quantities of dust. Until 1975, the Umicore zinc smelter was also using a similar process. After this date, it changed to an electrolytic process which considerably reduced its atmospheric emissions. Overall the dust fallout has affected an area of around 120 km2 where more than 55,000 inhabitants live. From 1994 to 2002, 10–15% of children (2–4 years old) living near Metaleurop Nord had blood lead levels exceeding 100 μg L−1 of blood (Declercq and Beaubois 2000; Leroyer et al. 2000). More recent measurements, taken after the smelter’s closure, showed that only 2.4% of children had blood lead levels over 100 μg L−1 (Declercq and Ladrière 2005). Despite this sanitary improvement, the mean blood lead level of these children still remains higher than the French average (42 versus 27 μg of lead per liter of blood) (Declercq and Ladrière 2004; Chatelot et al. 2008). Even if it is undeniable that, before 2003, the dust emissions coming from Metaleurop Nord were the main reason for the contamination of children, today it is contaminated soil that is the main source of pollutant exposure. Indeed many works have shown that direct ingestion of dust and soil particles via hand-to-mouth contact strongly contributes to exposure in children (Vostal et al. 1974; Duggan and Inskip 1985; Lanphear et al. 1998; Calabrese et al. 1999; Freeman et al. 2001; Johnson and Bretsch 2002).
Most current health risk assessments are based on total MTE concentrations. However these data do not always reflect the quantities that are available to the body (Brandon et al. 2006). The estimation of pollutant bioavailability is becoming more important in terms of health risk assessment and remediation efforts. Oral bioavailability is considered to be the product of bioaccessibility, absorption and metabolism (Oomen et al. 2003). The bioaccessible fraction is the fraction that is mobilized from its matrix (e.g. soil, dust, food, water) in the human gastrointestinal tract and becomes available for intestinal absorption (Oomen et al. 2006). The oral bioavailable fraction of a compound is the fraction that reaches the systemic circulation (i.e. blood stream) and can exert adverse effects. Implementation of the relative bioavailability factor in risk assessment is expected to lead to more realistic and less conservative estimation of exposure to a contaminant after soil ingestion (Oomen et al. 2006). In vitro testing regimes are used as predictors, as they do not provide absolute bioavailability data. This can only be done at present by in vivo techniques (Environment Agency and British Geological Survey 2002). However, in vivo tests are expensive, time-consuming and may involve animal experimentation, which should be avoided. This is why, since the 1990 s, in vitro techniques have been developed to assess the degree of MTE solubility in a simulated gastrointestinal environment imitating the leaching of a solid matrix. However, there are many differences in digestion characteristics between these in vitro methods that can lead to a wide range of bioaccessibility values (Oomen et al. 2002; van de Wiele et al. 2007). The Bioaccessibility Research Group of Europe (BARGE) undertook an international collaborative initiative to develop a unified method (the unified BARGE method, UBM) capable of providing reproducible, robust and defensible bioaccessibility data (Cave et al. 2006). The intent was to establish a standardized method that could provide a conservative estimate of the oral bioavailable fraction of pollutants from soil and could be used in human health risk assessment (Denys et al. 2007, 2009; Button et al. 2009). Moreover, the UBM test has just been validated against an in vivo model (young swine) for Cd, Pb and As (Caboche 2009).
The aim of this study was to (1) apply the UBM protocol in order to characterize the oral bioaccessibility of Cd, Pb and Zn from urban topsoils formerly contaminated by atmospheric emissions from two lead and zinc smelters and (2) contribute to the human risk assessment by using this in vitro tool. Specific attention was given to the physico-chemical soil parameters measured in order to understand how oral bioaccessibility is affected by soil characteristics. Several studies have assessed Pb bioavailable fractions in soils, but very few have looked at Cd and Zn bioaccessibility, although Cd is considered a carcinogen for humans (IARC 1987) and Zn may also be a problem for human health (Merian et al. 2004; Pierzynsky et al. 2005).
Materials and Methods
Topsoil (0–25 cm) was sampled from 15 ploughed horizons of kitchen gardens and from 12 lawns. Depending on the site surface, 5–13 elementary samples were taken using a 6-cm-diameter stainless-steel auger and pooled to make up a representative sample. Care was taken to avoid the proximity of pathways, vehicle tracks or roads. For lawns, the main roots were gently removed during soil sampling. Then, the soil samples were placed in plastic bags and brought back to the laboratory.
Topsoil Physico-Chemical Parameters
Physico-chemical parameters were determined at the INRA Soil Analysis Laboratory and are detailed in Douay et al. (2008). In brief, the soil samples were dried at a temperature below 40°C, crushed, and sieved to pass through a 2 mm mesh. Particle size distribution was obtained through sedimentation and sieving (NF X 31-107), pH was measured in a water suspension (NF ISO 10390) and organic carbon and total nitrogen were obtained by the NF ISO 10694 and NF ISO 13878 standards, respectively. Total carbonates were obtained by measuring the volume of CO2 released after a reaction with HCl (NF ISO 10693). Assimilated P (expressed in g P2O5 kg−1) was measured using an extraction by ammonium oxalate solution and spectrocolorimetric determination (NF X 31-161). Cationic exchange capacity (CEC) was obtained by percolation of 1 mol L−1 ammonium acetate solution, pH = 7 (NF X 31-130).
For each soil sample, a subsample was crushed to pass through a 250-μm mesh for total dissolution analysis. Calcination at 450°C and a mixture of hydrofluoric and perchloric acids, as described by the NF X 31-147 standard, were used for total dissolution of Cd, Pb and Zn. Manganese was extracted with ethylenediamine tetraacetic acid (EDTA) according to the NF X 31-120 standard. Free Fe and Al were extracted by a mix of solutions (0.267 mol L−1 sodium tricitrate, 0.111 mol L−1 sodium bicarbonate and 200 g L−1 sodium dithionite) using the protocol described by Mehra and Jackson (1960)), well known to dissolve the totality of iron oxides and oxyhydroxides. The Al, Fe, Mn, Zn and high concentrations of Cd and Pb were measured by inductively coupled plasma atomic emission spectrometry (ICPAES). Low concentrations of Cd and Pb were determined by inductively coupled plasma mass spectrometry (ICPMS).
All precautions were taken regarding protocol application and calibration. Quality control was based on the use of certified samples (BCR 141 and 142; GBW 07401, 07402, 07404, 07405 and 07406), samples from inter-laboratory comparisons, internal control samples and analysis duplicates.
Oral Bioaccessibility of Cd, Pb and Zn
For every four soils, which were replicated three times, a blank and a reference soil (SRM 2710) were used. The detection limits corresponding to the analysis of the gastric chyme were 0.08 mg kg−1 for Cd, 1.54 mg kg−1 for Pb and 0.06 mg kg−1 for Zn. The detection limits corresponding to the analysis of the gastro-intestinal chyme were 0.20 mg kg−1 for Cd, 3.99 mg kg−1 for Pb and 0.14 mg kg−1 for Zn.
MTE bioaccessibility is expressed as the ratio between the extracted concentration in gastric or gastro-intestinal phases and the total concentration before digestion.
To test the relationships between the oral bioaccessibility values and the soil total concentrations of Cd, Pb and Zn, linear regressions were performed. Furthermore, to investigate the principal factors determining soil bioaccessibility, a linear multiple regression was applied to correlate the measured gastro-intestinal bioaccessibility and some soil physico-chemical parameters (clay, assimilated P, pH, N, organic carbon, CEC, total CaCO3, Mn extracted with EDTA, free Al and Fe, total Cd, Pb and Zn concentrations). Before including one of these variables in the model, their linear relationship was tested. When no linearity was observed, the data were not included in the model. All statistical analyses were performed using STATISTICA 6.0 (Statsoft, Tulsa, USA).
Results and Discussion
Physico-Chemical Parameters of Studied Urban Soils
The soil description has been detailed by Douay et al. (2008), showing a high anthropogenic action with a thicker organo-mineral layer than their agricultural counterparts located in the same environment. The soils sometimes contained a variable load of coarse elements, for example gravels, slag and pieces of bricks and glass. The pedological development of these urban soils did not correspond to that which is commonly observed in regional soils (Douay et al. 2008). Moreover, in contrast to agricultural soils, the transition from the organo-mineral layer to the mineral one was often gradual, whereas it was distinct in agricultural soils, corresponding frequently to the bottom of the ploughed horizon.
The surface layers of the studied soils showed mainly a loamy texture. Depending on the site, however, two distinctions could be made. Around the Metaleurop Nord smelter, the soils contained higher clay contents, whereas around the Umicore smelter they showed higher sand contents. Soil pH was mainly neutral to slightly alkaline.
Mean, standard deviation (SD), minimum, maximum, first quartile (Q1), median and third quartile (Q3) values of physico-chemical parameters of the topsoils
g 100 g−1
g 100 g−1
Regarding the major pollutants (Cd, Pb and Zn), the lowest and highest concentrations were as follows: 3.1 and 31.4 mg kg−1 for Cd, 95 and 3,026 mg kg−1 for Pb and 326 and 6,908 mg kg−1 for Zn. The description of all studied elements in the urban topsoils has been detailed by Douay et al. (2008).
Oral Bioaccessibility of Cd, Pb and Zn in the Reference Soil
The reference soil, NIST SRM 2710, was used in order to validate the accuracy of the extraction protocol (n = 25) and the MTE analysis. The bioaccessibility values were significantly higher for the gastric phase than for the gastro-intestinal phase. For the gastric phase, the bioaccessibility values of Cd, Pb and Zn reached 58, 56 and 22%, respectively, of total content. In the gastro-intestinal phase, the bioaccessibility values of Cd, Pb and Zn reached 30, 27 and 11%, respectively, of total content. The repeatability of the test was good, as shown by the standard deviation. NIST 2710 was also employed by BARGE in an inter-laboratory study (Wragg et al. 2009). The affiliated laboratories that participated in this study were: the British Geological Survey (BGS), DHI Soil and Water (Denmark), Ohio State University (USA) and the RIVM. The results for Cd (in mg kg−1) were highly comparable. The BARGE inter-laboratory study obtained 14.8 ± 0.9 mg kg−1 (n = 4) for the gastric-only phase and 7.0 ± 1.6 mg kg−1 (n = 4) for the gastric and intestinal phase. In the present study, 12.6 ± 0.8 mg kg−1 (n = 25) was obtained for the gastric-only phase and 6.5 ± 0.9 mg kg−1 (n = 25) for the gastro-intestinal phase. The results for Pb were also comparable, even if the values in the gastric phase were slightly lower. The BARGE inter-laboratory study recorded 3,785 ± 445 mg kg−1 (n = 4) for the gastric phase and 1,138 ± 839 mg kg−1 (n = 4) for the gastro-intestinal phase, whereas in the present study 3,085 ± 371 mg kg−1 (n = 25) and 1,486 ± 259 mg kg−1 (n = 25) were obtained, respectively. The comparison with the BARGE inter-laboratory values suggested good reproducibility and accuracy of results using the UBM method. For Zn, there were no data published in the literature.
Moreover, the results obtained for Pb were compared with those of Denys et al. (2007). These authors used a previous version of UBM, i.e. end-over-end rotation for 2 h after addition of the gastric juice and for 2 h after addition of the duodenal and bile juices, whereas in the present study the rotation durations were 1 and 4 h, respectively. The Pb bioaccessibility values were 56 and 27% in this work versus 79% and 25%, in the gastric and gastro-intestinal phases, respectively. The variation observed in the gastric phase could be explained by the shorter rotation time.
Oral Bioaccessibility of Cd, Pb and Zn in the Contaminated Soils
Bioaccessible fractions of Cd, Pb and Zn expressed as a percentage of the total concentrations in the soils. Mean, standard deviation (SD), minimum and maximum are given for the 27 soils
The choice of bioaccessibility measure (i.e. gastric or intestinal compartment solubility) was a great issue for Cd, Pb and Zn, notably compared with others elements such as As and Sb. Indeed, As and Sb solubilities were shown to be similar in the gastric and intestinal compartments (Ellickson et al. 2001; Denys et al. 2009). However, in this present work, there was a sharp decrease in extracted Cd, Pb and Zn as they moved from the gastric fluid phase to the intestinal fluid phase of the sequential in vitro extraction. These results may be due to readsorption of metal onto the soil, complexation by pepsin or chemical precipitation of metals due to the higher pH environment of the intestinal compartment (Ellickson et al. 2001; Grøn and Andersen 2003).
Distribution of Cd, Pb and Zn within the 27 soil samples (expressed as percentage of total soil metal content) in the different fractions (exchangeable, reducible, oxidizable and residual form)
Relationships Between Concentrations Extracted with UBM, Total Contents and Selected Parameters of Soils
Relationships between predicted bioaccessible concentrations of Cd, Pb and Zn (expressed in mg kg−1) and soil parameters as given by stepwise regression models (significance level: P < 0.0001)
G bioacc Cd = 1.23 − 2.25 × [Fe] + 0.79 × [Cdtot] − 1.02E-03 × [Pbtot]
GI bioacc Cd = 0.68 − 0.55 × [Ntot] + 0.39 × [Cdtot]
G bioacc Pb = 171.7 − 1.09 × [CaCO3] − 225.9 × [Fe] + 0.68 × [Pbtot]
GI bioacc Pb = 1020.6 − 32.6 × [Ntot] − 131.1 × pH + 0.39 × [Pbtot]
G bioacc Zn = –54.5 − 0.41 × [Pbtot] + 0.80 × [Zntot]
GI bioacc Zn = 168.4 + 0.41 × [Zntot] − 1.88 × [Clay]
The negative correlations observed with the soil parameters such as pH, total nitrogen, clay, iron and carbonate concentrations provided evidence for a decrease of Cd, Pb and Zn bioaccessibility in soils with high values of these physico-chemical parameters. The MTE–soil component complexes seemed to be very stable. However they might dissolve during the gastric phase, involving in the intestinal fluid possible readsorption of MTE onto these soil components, reducing their bioaccessibility.
The observed relationships can be explained by the interactions of the metals with the different soil compartments. Indeed, it has been established that the major factors governing the distribution and dynamics of Cd, Pb and Zn in soils are pH, organic matter content, inorganic ligands, hydrous metal oxides, specific clay mineralogy and competition with other metal ions (Alloway 1995; Lothenbach et al. 1999; Davranche and Bollinger 2000; Violante et al. 2003; Abollino et al. 2006). Moreover, the bioaccessibility of Pb is known to be influenced by the metal speciation and the soil physico-chemical parameters, especially carbonates, as found in the multiple regression model. Some authors (Cotter-Howells and Thornton 1991; Davis et al. 1992) have suggested that Pb bioavailability was lower at mining sites compared with urban environments due to the existence of relatively insoluble forms of Pb associated with mining wastes (Gasser et al. 1996). These data explained very well the relationships obtained between Cd, Pb and Zn and pH, carbonates, clay and iron content. Concerning the correlations between Cd, Pb and total nitrogen observed in the gastro-intestinal phase, the high nitrogen content could be explained by the abundant use of nitrogen fertilizers on kitchen gardens and lawns. These observations provided evidence for a significant association of Cd and Pb to more stable forms, although not fully quantified in this present study.
Use of the In Vitro UBM Test for Human Risk Assessment
The UBM test has just been validated against an in vivo model (young swine) for Cd, Pb and As for the gastric and gastro-intestinal extraction phases (Caboche 2009). The results of this author showed very good correlations between the in vitro and the in vivo models. However, Caboche (2009) showed higher correlations between UBM gastric phase extraction and in vivo model than with gastro-intestinal phase extraction. Thus, for risk assessment, it may be more relevant to use the more conservative estimate, i.e. to take into account the values of the gastric phase. Indeed, in this study, significant differences were observed between bioaccessible values of the gastric and gastro-intestinal phases. The stomach bioaccessible concentrations (Cd, Pb and Zn) were very much higher than in the intestine.
In parallel with this, the present work showed that human bioaccessible fractions were largely related to total MTE concentrations as well as soil physico-chemical parameters and they were thus a better estimation of potential bioavailability. Therefore, the introduction of human bioaccessible concentrations gives more realistic indications for risk estimations. Recently Davis and Mirick (2006)) showed that estimates of soil ingestion rates in children were significant and, according to the considered tracer, mean soil ingestion rates ranged from 37 to 207 mg day−1. Using this input with our results, it was calculated that the bioaccessible Pb mass for a child living in the studied area, dependent on soil ingestion, could reach 35.5 μg per day. This value is given as an indication, as it do not give information about the health of people, taking into account neither the strong inter-individual variability of ingested quantities nor temporal variability. So, this work contributes to reduce the uncertainty of the assessment of people’s lead, cadmium and zinc exposure, as in Glorennec (2006)), by specifying the bioaccessible pollutant fraction.
In this study, the UBM method was used to simulate the human gastro-intestinal environment and to estimate the oral bioaccessibility of Cd, Pb and Zn in urban smelter-contaminated topsoils. Strong linear relationships were observed between oral bioaccessible values and total MTE concentrations. Moreover, human bioaccessible concentrations were also affected by physico-chemical soil parameters. Indeed, multiple regressions provided evidence for good interrelations between bioaccessible fractions and significant soil parameters such as pH, total nitrogen, carbonates, clay and iron content.
In these urban soils bioaccessibility showed wide variations, which can be explain by the variability of the physico-chemical soil parameters but also by the contamination route of these soils (dust deposition, anthropogenic addition of wastes, slag or ashes). The specificities of urban soils complicate the generalization of the approach. Despite this, this unified test revealed some valuable and accurate data for risk assessment.
This project represents the preliminary stage of a larger study with the aim of assessing metal bioaccessibility in different types of soils contaminated with different anthropogenic sources in the North of France.
The authors wish to thanks the Nord-Pas de Calais Council, the French Ministry of Research and the European Regional Development Fund (FEDER), for financial support. They thank Sébastien Denys, Karine Tack and Julien Caboche (Institut National de l’Environnement Industriel et des Risques) as well as the BARGE members (Bioaccessiblity Research Group of Europe) for scientific and technical help for the realization of the bioaccessibility tests.