JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 27 October 2006
Published in J Environ Qual 35:2066-2074 (2006)
DOI: 10.2134/jeq2005.0464
© 2006 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Clark, H. F.
Right arrow Articles by Erdil, R. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Clark, H. F.
Right arrow Articles by Erdil, R. M.
Agricola
Right arrow Articles by Clark, H. F.
Right arrow Articles by Erdil, R. M.
Related Collections
Right arrow Plant and Soil Interactions
Right arrow Heavy Metals
Right arrow Soil Chemistry

TECHNICAL REPORTS

Heavy Metals in the Environment

Sources, Sinks, and Exposure Pathways of Lead in Urban Garden Soil

Heather F. Clarka,*, Daniel J. Brabandera and Rachel M. Erdilb

a Geosciences Dep., Wellesley College, 21 Wellesley College Rd, Wellesley, MA 02481
b Wellesley College, 21 Wellesley College Rd, Wellesley, MA 02481

* Corresponding author (hclark2{at}wellesley.edu)

Received for publication December 6, 2005.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
The chemistry of Pb in urban soil must be understood in order to limit human exposure to Pb in soil and produce and to implement remediation schemes. In inner-city gardens where Pb contamination is prevalent and financial resources are limited, it is critical to identify the variables that control Pb bioavailability. Field-portable X-ray fluorescence was used to measure Pb in 103 urban gardens in Roxbury and Dorchester, MA, and 88% were found to contain Pb above the USEPA reportable limit of 400 µg g–1. Phosphorus, iron, loss on ignition, and pH data were collected, Pb-bearing phases were identified by X-ray diffraction, and Pb isotopes were measured using inductively coupled plasma mass spectrometry. Four test crops were grown both in situ and in Roxbury soil in a greenhouse, and plant tissue was analyzed for Pb uptake by polarized energy-dispersive X-ray fluorescence. Variation at the neighborhood scale in soil mineralogical and chemical characteristics suggests that the bioavailable fraction of Pb in gardens is site specific. Based on Pb isotope analysis, two historical Pb sources appear to dominate the inventory of Pb in Roxbury gardens: leaded gasoline (207 Pb/206 Pb = 0.827) and Pb-based paint (207Pb/206 Pb = 0.867). Nearly 70% of the samples analyzed can be isotopically described by mixing these two end members, with Pb-based paint contributing 40 to 80% of the mass balance. A simplified urban human exposure model suggests that the consumption of produce from urban gardens is equivalent to approximately 10 to 25% of children's daily exposure from tap water. Furthermore, analysis of over 60 samples of plant tissue from the four test species suggests that in these urban gardens unamended phytoremediation is an inadequate tool for decreasing soil Pb.

Abbreviations: XRF, X-ray fluorescence • FP-XRF, field-portable X-ray fluorescence • PED-XRF, polarized energy-dispersive X-ray fluorescence • LOI, loss on ignition • ICP–MS, inductively coupled plasma mass spectrometry • BLL, blood lead level • XRD, X-ray diffraction • Pbsoil, soil lead concentration • Pbplant, plant lead concentration • wt %, weight percent • IEUBK, Integrated Exposure Uptake Biokinetic


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
LEAD CONTAMINATION OF SOILS is a serious public health issue that is compounded when urban gardening is an integral part of community life and food security. Lead poses a particular threat to children because of its properties as a neurotoxin that inhibits development (Xintaras, 1992; Agency for Toxic Substances and Disease Registry (ATSDR), 2000). Many urban environments are polluted by historic and current use of Pb-based products, such as Pb-based paint and leaded gasoline, with peak usage in 1925 and 1979, respectively; both sources were phased out in the USA by the end of the 1970s (Rabinowitz, 1987; Filippelli et al., 2005). Identifying the sources of Pb in this setting is essential because source is one of the factors that controls the mineralogy of the Pb present, which in turn is a major factor of bioavailability. In urban areas, soil is a key sink for Pb in the environment and a major site for human exposure (Mielke and Reagan, 1998; Shinn et al., 1998). Lead in soil can enter the human body through a variety of pathways, primarily through the ingestion or inhalation of soil dust, and less significantly through the consumption of produce grown in contaminated soil (USEPA, 1997; Chaney et al., 1997; Hettiarachchi and Pierzynski, 2002). Lead abatement and exposure reduction programs have been effective in the USA, resulting in an 80% decrease in the number of children nationwide with elevated blood lead levels (BLL) since the 1970s. However, whereas the national average of children with BLLs above the Center for Disease Control safe level of 10 µg dL–1 is now only 2.2%, the urban average remains at 15% (Agency for Toxic Substances and Disease Registry, 2000).

Exposure to Pb from soil is an environmental justice issue because it disproportionately affects urban, poor, and minority communities. The Boston area communities of Roxbury and Dorchester, MA, illustrate the demographic distribution of Pb poisoning. Dorchester has the largest African-American and Cape Verdean population in the Boston area, the largest percentage of houses built before 1940, a high rate of poverty with 23% of the population living below the poverty line, and the highest percentage of children with elevated BLLs, with North Dorchester's BLLs ranking 50% higher than the Boston average (Dorchester Environmental Health Coalition, 2003; Boston Public Health Commission, 2004). Roxbury and Dorchester are also unique settings because backyard gardening is an important element of community life and food security. This is especially evident among the Cape Verdean population, which relies on backyard gardening to provide produce that is a central element of a traditional diet.

The research approach of this study employs novel methods in environmental science to determine the source, mobility, and bioavailability of urban Pb. Typically, soil Pb concentration (Pbsoil) and Pb concentration absorbed by plants (Pbplant) are measured by atomic absorption spectrometry or inductively coupled plasma mass spectrometry (ICP–MS) and require lengthy and expensive digestions (Clevenger et al., 1991; Blaylock et al., 1997; Ryan et al., 2001). These techniques limit the number of samples that can be processed, thereby making large scale, in situ experiments difficult. Spittler and Feder (1979) and Litt et al. (2002) initiated the use of new testing methods for Pbsoil by using X-ray fluorescence (XRF) technology to analyze a large number of samples in situ in the Boston area. A rapid assessment protocol using field-portable X-ray fluorescence (FP-XRF) technology is extremely cost effective and quickly identifies areas (i.e., hot spots) where more thorough sampling and studying are necessary.

The other critical component of this research model, which is also used by Litt et al. (2002), is the partnership with community organizations. Collaboration with grassroots organizations enables scientific research to address problems influencing a community and to effectively communicate results; in this study, a partnership with The Food Project, a grassroots organization committed to promoting urban agriculture and educating the Roxbury and Dorchester communities on the threats posed by Pb, provided access to all of the backyard gardens tested.

Since the residents of the Roxbury and Dorchester areas rely so heavily on homegrown produce, the urban Pb cycle and the role of plants as potential environmental sinks, pathways of exposure, and remediation tools must be clearly understood (Fig. 1). The principal objectives of this research are to (1) devise a rapid screening protocol for Pbsoil, (2) assess the spatial variation of Pbsoil both on the neighborhood scale and in individual gardens, (3) evaluate the chemical and physical properties of Pb in soil as they influence bioavailability, (4) explore the potential of various plant species as phytoremediation tools, (5) fingerprint the sources of Pb in the environment, and (6) quantify the human exposure to Pbplant through the consumption of produce grown in contaminated urban soil.


Figure 1
View larger version (22K):
[in this window]
[in a new window]
  		Fig. 1. The environmental Pb cycle is unique in the urban environment. This flow chart, modified from Litt et al. (2002), includes urban produce, which has not been quantified as a sink or a pathway of exposure.

 

   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Soil and Plant Sampling
Soil was collected from 103 backyard gardens between 2003 and 2005. The Food Project enlisted gardeners to participate in this study; the only criteria for screening gardens for Pb was that they were actively used for growing produce. A minimum of four samples, generally two from the surface horizon (0 to 10 cm) and two from the rooting depth (30 to 40 cm), were collected from each site to produce a representative profile of Pbsoil. The average garden was 20 m2, and approximately 50 g of soil were collected in plastic bags from distinct areas of each garden. Aliquots of 4 g of each soil sample were dried at 50°C for 2 d and placed in XRF sample cups with 6-µm-thick Mylar film windows for analysis.

Nine gardens were used as plots to test phytoremediation, to measure additional soil chemical properties, and to size-fractionate (sieve) soil to observe the variation in Pbsoil as a function of grain size (from 4 to 0.034 mm). These nine gardens were selected based on permission from residents, as well as the wide range of Pbsoil in these plots (475 to 3600 µg g–1). Out of these nine plots, five were located in the back of the house (therefore separated from the roadside by the structure), and four were located in the front of the house (separated from the roadside only by the sidewalk and a wooden or chain link fence). All gardens ran adjacent to the residential structure.

The plant species grown in this study included the documented hyper-accumulating species of mizuna mustard (Brassica rapa), collards (Brassica oleracea), and sunflowers (Helianthus annuus), as well as beans (Phaseolous vulgaris) (Prasad, 2003). The selection criteria for these species were based on the plant's ability to evapotranspire soil water, bioaccumulate contaminants, mature quickly, and on the frequency of use by neighborhood gardeners (Chaney et al., 1997). Plants were individually hand washed, soaked in dionized water for 10 min, and rinsed again. Concentrations of Si and Al were monitored by polarized energy-dispersive X-ray fluorescence (PED-XRF) to ensure the removal of all surface soil and dust before testing.

X-ray Fluorescence
Soil lead concentration was analyzed by two complementary XRF approaches. An FP-XRF Niton (NITON XLi 700 Series Environmental Analyzer, Thermo Electron Corporation, Billerica, MA) was used for in situ measurements as well as for prepared samples, and a Spectro XEPOS (Spectro Analytical, Kleve, Germany) PED-XRF instrument was used to analyze a subset of samples (10% randomly selected). This additional testing was conducted in accordance with USEPA method 6200, which, for quality control purposes, requires that a subset of in situ FP-XRF samples be analyzed by a complementary analytical method. Using a 12-mci 109Cd source, FP-XRF achieved ± 10% analytical error for Pbsoil by counting for 60 decay-corrected seconds.

Plant lead concentration was analyzed by PED-XRF to achieve an error of ± 5%. Preparation for analysis of plant samples involved drying plants to constant mass at 50°C, grinding plant tissue for 5 min in a tungsten carbide mixer mill, combining with SpectroBlend binding agent, and pressing the material into a pellet under 10 metric tons of pressure. Additionally, total soil P and Fe were measured via analysis of soil pellets, created as stated above, using PED-XRF. Testing of all unknown samples via XRF was bracketed with the National Institute of Standards and Technology 2709 or 2711 Standard Reference Material. Concentrations of all elements of interest in the standards remained within ± 10% of accepted values.

Soil Chemical Properties
Soil pH was measured using a 1:2 mass ratio of soil and deionized water (10 g/20 mL). Loss on ignition (LOI) was measured as a proxy for total organic matter by placing approximately 5 g of soil in a furnace at 550°C for 4 h and then measuring the mass difference (Pichtel et al., 2001). Preliminary X-ray diffraction (XRD) work to identify the mineral phases present was conducted on four soil samples that had been sieved (<400 µm) and density-separated using sodium polytungstate with a density of 2.89 g cm–3 to enrich Pb to detectable limits. These four samples (light and heavy mineral fractions from two locations) were selected because they represent the two ends of the range of Pb isotope values. Jade software (Materials Data, Livermore, California) with Search/Match of the FIZ-Inorganic Crystal Structure Database minerals database and Rietveld whole pattern fitting (which involves the simultaneous refinement of the amounts and crystallographic parameters in mixtures of minerals) were used to identify the Pb-bearing phases (Rietveld, 1969; Synder and Bish, 1989). Analyses were conducted using a rotating Cu anode RU 300 generator (Rigaku, Tokyo, Japan).

Lead stable isotope (206Pb, 207Pb, and 208Pb) analyses were conducted at the analytical geochemistry facilities in the Department of Earth Sciences at Boston University using a Jobin Yvon Ultima ICP–MS (Analis, Suarlée, Belgium). The observed analytical precision was generally less than 0.5% (standard error) and the mean (n = 12) 207Pb/206Pb isotopic ratio measured on USGS Pb isotope standard BHVO-2 was 0.833 ± 0.004. Given the standard deviation typically obtainable with ICP–MS, these results compare with Li and Niu's (2003) and Elburg et al.'s (2005) examinations of BHVO-2 using multi-collector ICP–MS instrumentation. Their estimates for the 207Pb/206Pb ratio for BHVO-2 range from 0.8325 to 0.8331. Sample preparation procedures for ICP–MS analysis followed a general microwave digestion protocol; 50 mg of powdered sample was digested in 6 mL of HNO3, 2 mL of HCl, and 2 mL of HF, and microwaved at 210°C. Ten mL of 5% H2BO3 and 500 µL of H2O2 were added before the vessels were microwaved again at 150°C. Samples were then diluted a total of 6000 times and sonicated for 30 min to ensure total dissolution. Complete sample preparation technique was followed according to Plank (2000).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSION
 REFERENCES
 
Soil Homogeneity
The average Pbsoil for the 843 individual soil samples measured in this urban neighborhood is 950 µg g–1, which is more than twice the EPA reportable limit of 400 µg g–1. The Roxbury and Dorchester plots tested have been gardened for 2 to 5 yr on average. As a result, the Pbsoil was found to be homogeneous in depth and surface distribution, and hot spots were not observed because of years of tilling the soil for gardening. Therefore, it is feasible to estimate the average Pbsoil in each garden from just 4 to 6 discrete samples. Table 1 shows average Pbsoil values based on a large number of FP-XRF in situ measurements as compared to a smaller, discrete sample set from the same garden. The resulting means and standard deviations from the two sampling and analytical protocols indicate that there is no statistically significant difference between discrete and in situ sampling and that the collection of approximately 5 samples per garden provides an adequate estimate of average Pbsoil.


View this table:
[in this window]
[in a new window]
  		Table 1. Soil homogeneity and method verification.

 
Regional Lead Contamination
Soil lead concentration across the Roxbury and Dorchester neighborhoods spans a wide range of values. Figure 2 illustrates the spatial and statistical distribution of Pbsoil across 103 backyard gardens (n = 843) as well as the location of the nine phytoremediation test plots within approximately ten city blocks. In this study, 88% of the gardens tested contained Pbsoil above the USEPA reportable limit of 400 µg g–1. Forty-nine percent of the gardens had Pbsoil between 400 and 1200 µg g–1. A 2002 study by Litt et al. in the Roxbury and Dorchester areas, also conducted using FP-XRF, found comparable concentrations, with 55% of sites tested containing Pbsoil between 400 and 1200 µg g–1. Litt et al. (2002), however, found Pbsoil above 5000 µg g–1 in 3% of yard soil, which was not observed in this study. An explanation for this difference is that the study tested uncultivated backyard lots as opposed to gardens with well-mixed soil and obtained samples from along the house drip lines. The elevated Pbsoil and hot spots found in Litt et al. reflect Pb-based paint as the dominant source of Pb (determined by proximity to residence), whereas the random distribution of Pbsoil in gardens found in this study suggests that the garden soil is a more well-mixed matrix.


Figure 2
View larger version (23K):
[in this window]
[in a new window]
  		Fig. 2. Regional Pb contamination has been measured by field-portable X-ray fluorescence (FP-XRF) and the distribution of Pb can be illustrated by (a) GIS mapping and (b) a histogram of frequencies of observed soil Pb concentrations (Pbsoil).

 
Soil Mineralogy
In addition to soil chemical properties, the mineralogical phases associated with Pb in soil plays an important role in Pb bioavailability. In the gardens tested in this study, the average soil is characterized by 27% gravel, 68% sand, and 5% clay and/or mud. The size fraction of 0.5 to 0.25 mm represents approximately 30% of the soil by mass and accounts for 32% of the Pb. The soil in the size fractions <0.044 mm represent only 5% of the bulk soil by mass but a disproportionate amount of the Pb, 10%, is contained in these size fractions. Since the combustion of leaded gasoline results in the deposition of fine particulate matter, it is likely that the Pb measured in these size fractions represents the input from this source. Fine particles, specifically particulate matter <100 µm, are often considered wind-transferable (Watson and Chow, 2000). The particles that are redistributed by wind–both into other gardens and into homes–represent the most significant exposure threat via inhalation or ingestion. Therefore, understanding the bioavailability of this size fraction in particular is necessary in further remediation work.

Garden soil measured via XRD provides a representative profile of general soil mineralogy. Soil from the two garden plots (plots 1 and 4) that were density-separated yielded similar mineralogy in the light or float fraction (density < 2.89 g cm–3). As expected, quartz represented 35 to 40 wt% and sodium and potassium feldspars accounted for 35 to 43 wt% with the remainder of this fraction composed of clay minerals (illitic in composition) and chlorite group minerals. The heavy soil fractions (density > 2.89 g cm–3) yielded a complex mineral assemblage dominated by iron-bearing chain silicates (hornblende and augite). These phases comprised 48 to 55 wt% of the fraction. The stability of Pb in soil is influenced by the significant finding of 6 to 11 wt% of manganese and iron oxides, which are minerals that often contain surfaces that can act as binding sites for Pb.

The specific mineral phase and the degree of crystallinity associated with Pb in soil are also determining factors in the bioavailability of Pb. For example, highly crystalline oxide-bearing Pb phases will be less susceptible to changing soil chemical characteristics than carbonate Pb-bearing phases (e.g., basic lead carbonate, white lead, or hydrocerussite 2PbCO3·Pb(OH)2), which formed the white pigment for lead paints (Hall and Tinklenberg, 2003). In the heavy soil fractions (density > 2.89 g cm–3), phoenicochroite [Pb2O (CrO4)] was identified in the sample with higher bulk Pb (plot 4) at an abundance estimated to range from 0.5 to 1.0 wt %. These XRD results are further supported by PED-XRF results that identify a correlation between Pb and Pb/Cr (R2 = 0.86) in soil samples. This lead chromate mineral has been previously observed in contaminated soils and is specifically associated with paint manufacturing (Treiman, 2000). The solubility of this phase is controlled by pH and at high pH (>8) phoenicochroite dissolves in favor of the HPbO2– ion. The remainder of the bulk soil Pb in the samples tested is either present in phases where abundance is less than 0.5 wt % (the working limit of detection for many Pb-bearing phases in this matrix) or is poorly crystalline or amorphous. This observation suggests that two distinct geochemical sources of Pb are present and points to Pb-based paints as an important source. Additional research utilizing electron microprobe analysis, sequential chemical extraction, or X-ray absorption spectroscopy is required to more precisely identify all bonding environments of Pb in the soil and to assess the relative bioavailability of each reservoir of soil Pb.

Variables Controlling Lead Bioavailability
Lead mobility and bioavailability are largely controlled by several key soil chemical characteristics including P, Fe, LOI, and pH. Iron hydroxides [such as goethite, Fe(OH)3] and total organic matter (estimated by measuring LOI as a proxy) create anionic surface bonding sites for Pb cations, and dissolved PO4 leads to Pb phosphate precipitation from soil solution, also decreasing the mobility of Pb in soil (Appel and Ma, 2002; Brown and Chaney, 2003; Ryan et al., 2004). Similarly, increased pH decreases Pb mobility and bioavailability since less H+ ions are available to compete with Pb cations for binding sites (Hettiarachchi and Pierzynski, 2004). The nine test plots in this study exhibit a wide range of soil chemical characteristics (Table 2). For example, P varies by a factor of five whereas pH varies by almost two pH units. This variability suggests that multiple Pb sequestration mechanisms are operating in these gardens, further complicating the design of effective remediation.


View this table:
[in this window]
[in a new window]
  		Table 2. Soil chemical characteristics and plant uptake.

 
Empirical Phytoavailability
Relative Pb bioavailability has been measured qualitatively as a function of plant uptake by several species. Table 2 shows Pbplant by species and plot. The average mustard Pbplant was measured at 46 ± 30 µg g–1 (n = 58). This sample set was harvested from two distinct growing locations: 42 samples grown in a greenhouse showed an average Pbplant of 39 ± 22 µg g–1 and 16 grown in situ contained an average Pbplant of 56 ± 32 µg g–1. Despite greenhouse mustards not growing to full maturity, the greenhouse appears to be an appropriate setting for testing the bioavailability of Pb in urban garden soils. In this controlled environment, it will be possible to isolate the soil chemical characteristics that play a key role in determining the bioavailability of Pb in this soil matrix.

Figure 3 illustrates the correlation between Pbsoil and Pbplant in mustards. The correlation between initial Pbsoil and uptake (R2 = 0.85) indicates that bulk Pbsoil is one of the controlling factors in uptake and exposure. The impact of other soil chemical variables on bioavailability could not be determined from this study since no other systematic correlations were observed between Pbplant and soil chemical properties.


Figure 3
View larger version (13K):
[in this window]
[in a new window]
  		Fig. 3. The Pbsoil and plant Pb uptake (Pbplant) in Brassica rapa (mustard) tissue analyzed by XRF. Mustards were grown in both the greenhouse (n = 4) and in situ in the test plots (n = 8). Each point represents an average of 2 to 4 Pbsoil measurements and 5 to 8 Pbplant measurements.

 
Mustard plants are the primary data presented because that species was grown and harvested with the greatest success and in the largest amounts, but collards, beans, and sunflowers were also tested. The average Pbplant in the foliar tissue of bean plants was 14 ± 5 µg g–1 (n = 8), whereas the Pbplant in the bean pod was 2 ± 0.7 µg g–1 (n = 6). Sunflowers and collard harvests were poor, but preliminary results (n = 4) indicate that sunflowers accumulate approximately 47 ± 9 µg g–1 and collards accumulate approximately 14 ± 3 µg g–1. A subset of samples were also analyzed by sections, specifically foliar tissue and root tissue; all plant species were found to have root Pbplant approximately three times greater than Pbplant in foliar tissue. For mustard plants, Pbplant in roots was measured at 115 µg g–1; this distinction is important when estimating exposure via ingestion of produce, and only foliar Pbplant values are used in exposure measurements in this study.

Overall, the low levels of Pb accumulation in the species studied indicate that unamended phytoremediation (phytoextraction relying solely on the natural ability of the plants to uptake metals without adding any soil amendments to alter the soil chemistry) is not an adequate remediation tool for the Roxbury and Dorchester neighborhoods.

Sources of Lead
To differentiate the relative contribution of the two Pb sources (Pb-based paint and leaded gasoline additives) to urban garden soil, both trace elemental ratio analysis (Adgate et al., 1998) and Pb isotopic analysis (Rabinowitz, 1995; Gwiazda and Smith, 2000; Brabander et al., unpublished data, 2006) were conducted on a subset of soil samples. Whereas white Pb (basic Pb carbonate) was by far the dominant Pb-based white pigment used in paints (Rabinowitz and Hall, 2002), other pigments were developed to make paint more resistant to weathering. These pigments included basic Pb sulfate (PbO·PbSO4), leaded zinc oxide (ZnO + PbSO4), and Pb titanate (PbTiO3). After the ban of Pb-based paints in the USA in 1978, titanium dioxide (TiO2) replaced the Pb-derived paint pigments (Hall and Tinklenberg, 2003). Therefore, elevated Ti in garden soils can be considered an indicator of the presence of total paint products in soils. The Ti was measured by PED-XRF (n = 30) and a correlation exists (R2 = 0.87) between Pbsoil and Pb/Ti, indicating that Pb-based paint is a contributor to the total soil Pb burden (Fig. 4a).


Figure 4
View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. Sources of Pb were identified using (a) trace element ratio and (b) isotopic ratio (measured by inductively coupled plasma mass spectrometry (ICP–MS)) analysis techniques. Both graphs represent data (n = 4) from a nonurban residence with a known and isolated Pb-based paint Pb source. The data in (b) is compared to the isotope ratio range for Pb-based paint and a leaded gasoline signature measured by Rabinowitz (1986).

 
A comparison of the ratios of 207Pb/206 Pb to the ratio of Pb/Ti can be used to further distinguish between the primary Pb sources. Two distinct clusters of data points can be observed in Fig. 4b; one group characterized by a Pb/Ti ratio less than 1.0 and a high 207Pb/206Pb that is associated with Pb-based paint; this cluster includes the four data points collected from a nonurban residence that was undergoing Pb-based paint remediation at the time of sample collection and is geographically isolated from other major inputs of environmental Pb. The other cluster is characterized by a Pb/Ti ratio greater than 1.0 and a low 207Pb/206Pb that is associated with leaded gasoline. Figures 4b and 5 contain Pb isotope measurements obtained by Rabinowitz (1986) that represent the signature range of values for Pb-based paint and leaded gasoline from accumulated fallout in the roadside soil matrix in the Boston area. The isotope data from garden soils clearly suggest that these two Pb sources play a significant role in the Pb contamination of gardens in the Roxbury and Dorchester areas. It is important to note that both Pb-based paint and leaded gasoline have varied Pb isotopic compositions because of the wide variety of manufacturers of Pb-based paints and the numerous locations of Pb ore bodies used in manufacturing of tetraethyl Pb gasoline additive (Hurst, 2000; Rabinowitz, 2005).


Figure 5
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 5. Isotope/isotope ratios of size-fractioned (4 to 0.037 mm) samples measured by inductively coupled plasma mass spectrometry (ICP–MS) from four of the urban phytoremediation test plots and the nonurban residence with an isolated Pb-based paint source. Error bars represent analytical uncertainty. The proposed end-member isotopic compositions for Pb-based paint and leaded gasoline are from Rabinowitz (1986). The plotted symbols represent the mean and the shaded ellipses represent the standard deviation. Ayuso et al. (2004) represents the atmospheric Pb composition in the USA in the 1970s and Mississippi Valley Type ore.

 
Isotope/isotope ratio plots are commonly used to examine mixing and mass balance relationships between multiple Pb sources. Table 3 illustrates that isotopic variation exists within a single garden and is differentiated by grain size. It appears that the finest grain sizes and highest Pbsoil are most associated with the isotopic signature of leaded gasoline. A second key observation is that there is variation in the isotopic composition between gardens, suggesting that spatially controlled variables are affecting the relative contribution of historic Pb sources. The Pb isotopic composition of the garden soils is defined by a mixing trend bound by the two dominant historic sources. In Fig. 5, data from Rabinowitz (1986) is used to constrain the likely end member isotopic compositions of Pb-based paint and Pb derived from the combustion of leaded gasoline, whereas the data presented from Ayuso et al. (2004; and references therein) provides the average fallout Pb isotopic composition for the USA during the 1970s.


View this table:
[in this window]
[in a new window]
  		Table 3. Soil Pb isotopic fingerprint by grain size.

 
Nearly 70% of the samples analyzed can be adequately described by a two-component mixing model using Rabinowitz's (1986) time-integrated and regionally dependent leaded gasoline source composition as gauged from urban roadside soil and Pb-based paint analyses. However, garden plot 4 appears to fall off this trend. Rabinowitz (2005) has recently documented that there can be significant isotopic variation within an ascribed Pb source (e.g., smelter, paint, and leaded gasoline). A range of isotope values results from both the spread of geological ages within Pb ore bodies and variation in the ore bodies used by a single manufacturer or point-source. The observed isotopic compositions in plot 4 may therefore reflect the use of an isotopically distinct batch of Pb; however, the 207Pb/206Pb ratios are not distinct enough to correlate with a specific or unique manufacturer or ore body.

Assessing the mass balance of anthropogenic sources of Pb to soil is critical to the design of site-specific remediation because Pb speciation, in part, controls bioavailability and remediation options. The mass balance and relative contribution of the two sources can be determined using a two-component mixing model that quantifies the fraction (F) of each source with the following equation:

Formula 1[1]
where 207Pb/206Pbgas = 0.827 and 207Pb/206Pbpaint = 0.867 (these are mean values from Rabinowitz (1986) expressed as the center of the ellipses/bars (Fig. 4b and 5) whose boundaries are defined by the standard deviation of the sample set). Using this equation, the Fpaint for plot 1 = 40% and the Fpaint for nonurban soil (collected during a Pb-based paint remediation project from a residence on the suburban Wellesley College campus) is 86%. This shows that there is a variable range in the relative contribution of these two end-member sources of Pb. It is worth noting that given the wide range of isotopic compositions observed in the proposed end members, it is possible to construct numerous mixing line trends. Additional soil and plant isotope analyses are planned to determine whether additional Pb sources (e.g., local smelters and scrap dealers) are required to better model the observed isotopic variation and to evaluate the relative bioavailability of each Pb source.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSION
 REFERENCES
 
The bioavailability of Pb in soil has two important implications: the impact on human exposure to Pb and the potential of various remediation schemes. The role of diet as a pathway of exposure to the human system, particularly produce grown in contaminated soil, is not well quantified. To put the relative exposure risk of produce in context, a range of exposure possibilities can be estimated using the uptake data from this study. Following a format presented by Hemond and Solo-Gabriele (2004), an estimate of exposure can be made with the equation:

Formula 2[2]
where Pbplant is the concentration of Pb in the edible portion of plant, IR is the ingestion rate, A is the percent absorbed, B is the percent bioavailable, and f is the fraction of days per year the produce is ingested. An estimate for dose of ingestion can be made with the following figures:

Pbplant = 20 to 40 µg g–1. This is the mean for mustard, collard, and bean edible tissue. To make this calculation as realistic as possible, a separate set of plants that were analyzed for Pbplant by PED-XRF were washed in a manner to replicate kitchen-style washing, which leaves some residual dust on the surface of the plant and increases Pbplant value.
IR = 0.00386 g kg–1 body weight per day (dry weight). This is the sum of the average consumption of mustards, collards, and beans for all demographic groups (USEPA, 1997). Additionally, for calculation purposes it is needed to assume that average weight of a 6 to 7 yr old child is 19 to 32 kg (Center for Disease Control and Prevention, 2000).
A = 50%. This is the default value given by the USEPA (2005). It is likely that this may be an overestimate, and values as low as 20% (Maddaloni et al., 1998; Agency for Toxic Substances and Disease Registry, 2000) have been suggested. For this calculation, this parameter indicates the average amount of Pb that moves from the site of administration to the systemic circulation; the difference is excreted.
B = 30% (Hettiarachchi and Pierzynski, 2004). For this calculation, this variable is a separate parameter that is specific to the chemical speciation of Pb and could therefore vary. This variable represents the fraction of the Pb that gains access to the site of action once it has entered the systemic circulation.
f = 40% (USEPA, 1997).

Using these values, the range of exposure for children between 6 and 7 yr from Pbplant of 0.09 to 0.33 µg per day can be estimated.

Although blood generally carries only a small fraction of the total Pb body burden, it serves as the initial receptacle of absorbed Pb and distributes Pb throughout the body making it available for other tissues or excretion, and BLL serves as the measure of this exposure (Agency for Toxic Substances and Disease Registry, 2000). The relative contribution of different sources of Pb to BLL can be illustrated with Eq. [3]:

Formula 3[3]

To put the Pbplant exposure in context, it can be compared with Pbdrinking water. The USEPA has developed an Integrated Exposure Uptake Biokinetic (IEUBK) model for evaluating Pb exposure in children. Using the IEUBK default parameters for Pb absorption (50%) and the Pb in tap water (4 µg L1), and the estimate by Ershow and Cantor (1989) of 6- to 7-yr-old children's consumption of tap water (0.65 L per day), Eq. [2] can be used to estimate that drinking water accounts for 1.3 µg of children's daily exposure. Comparing this value to the ranges of exposure from ingestion of garden produce, Pbplant accounts for approximately 10 to 25% of children's exposure from tap water. Although the contribution from direct ingestion of extremely elevated Pbsoil observed in the <37-µm size fraction of garden soil (Table 3) is likely the primary pathway of exposure, this work demonstrates that exposure from produce is both a quantifiable and a nontrivial component of the total Pb exposure pathway. Additional work to further size-fractionate soil and evaluate mass balance, along with additional refinements in produce intake models will allow for a more complete exposure assessment.

While children are the most at-risk population for Pb poisoning, estimating exposure in adults is critical as well, specifically for the vulnerable population of pregnant women. The placenta has been shown to be a poor barrier to Pb, and the relationship between maternal BLL and umbilical cord BLL has been demonstrated to be linear across a wide range of BLLs (Graziano et al., 1990). The calculations above represent the first step in estimating the total and mass balance of Pb exposure.

Remediation options are the second area that is influenced by the bioavailability of Pb. For The Food Project, which is committed to an organic philosophy and is operating on a limited budget, the use of synthetic chelating agents that affect the mobility of Pb and its uptake by plants (i.e., Pb phytoextraction) is not an economic or philosophic remedial option. In this urban community the remediation scheme needs to be cost-effective, in situ, and low-skill, so that residents can implement it themselves. Currently, The Food Project builds raised beds and supplies compost for residents, and recommends growing leafy-green, high-accumulating species in the raised bed and growing fruiting plants that do not accumulate Pb in the edible portion in the contaminated soil. However, long-term remediation or stabilization techniques are desired. The use of soil amendments such as phosphate material or clay minerals are all of interest to the researchers of this study and to The Food Project with the goal of decreasing Pb bioavailability to keep backyard gardening a safe and healthy practice (Farfel et al., 2004).


   CONCLUSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
REFERENCES
 
Results from this study indicate that anthropogenic Pb has a significant presence in the urban environment and that Pb residence time in soil makes it a persistent pollutant originating from several sources. The wide range of soil chemistry observed in this study demonstrates that remediation schemes will need to be site-specific to be effective (Table 2). Furthermore, produce grown in urban gardens acts as a quantifiable pathway of exposure to humans, but the concentrations and percent contribution as compared with drinking water suggests that backyard gardening can continue as an important element of community life. The research methods utilized for this study, of academic and nonprofit partnerships as well as the rapid screening protocols utilizing FP-XRF, allow environmental science research to occur on a large scale and improve the ability to share helpful information with communities. Urban areas, in particular, suffer from a disproportionate burden of Pb in the environment and in children's blood. With this in mind, lead poisoning prevention and remediation programs must be designed to target specific demographic needs and environmental justice issues.


   ACKNOWLEDGMENTS
 
All work was supported by grants from the Howard Hughes Medical Institute, the Brachman-Hoffman Fund, and Wellesley College. This study would not have been possible without the partnership with The Food Project and their commitment to urban education, environmental justice, and sustainable urban agriculture. Special thanks to the editor William Berti and the four anonymous peer reviewers of the paper. Additional thanks to James Besancon, Geosciences Department, Wellesley College, and Peter Kloumann, Department of Material Sciences, MIT, for XRD assistance, and to Louise Bolge and Andrew Kurtz, Department of Earth Sciences, Boston University, for ICP-MS assistance.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Clark, H. F.
Right arrow Articles by Erdil, R. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Clark, H. F.
Right arrow Articles by Erdil, R. M.
Agricola
Right arrow Articles by Clark, H. F.
Right arrow Articles by Erdil, R. M.
Related Collections
Right arrow Plant and Soil Interactions
Right arrow Heavy Metals
Right arrow Soil Chemistry

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome