Chaney, R.L., H.W. Mielke and S.B. Sterrett. 1989. Speciation, mobility, and bioavailability of Soil Lead. [Proc. Intern. Conf. Lead in Soils: Issues and Guidelines. B.E. Davies & B.G. Wixson (eds.)]. Environ. Geochem. Health 11(Supplement):105 129.
Speciation, Mobility, and Bioavailability of Soil Lead
Rufus L. Chaney
USDA-Agricultural Research Service, Soil-Microbial Systems Laboratory, Building 318, BARC-West, Beltsville, Maryland, 20705, USA.
Eastern Shore Agricultural Experiment Station, Virginia Polytechnic Institute and State University, Painter, Virginia, 23420, USA
ABSTRACT The contamination of urban soils by Pb is a persistent source of risk to children. Pb concentration increases as soil particle size decreases, and smaller particles are more easily moved into homes to become Pb-rich housedust, and onto hands where mouthing behaviour allows children to ingest soil particles. We have evaluated the roles that chemical speciation of soil Pb and bioavailability of Pb in ingested soils play in risk of soil Pb to children. Nearly all forms of Pb pollution [automotive PbClBr and PbSO4; mining PbS and PbCO3; paint PbCO3Pb(OH)2 and PbCrO4] are transformed by soil chemical and microbiological processes into adsorbed soil Pb. Soils have a high capacity to adsorb Pb on to surfaces of hydrous Fe oxide and organic matter. Typical soils can hold over 10,000 mg Pb/kg in short term adsorption experiments, and larger amounts of Pb in longer studies. Crystalline PbSO4 added to a moist soil was quickly transformed to non-crystalline materials. A fundamental property of Pb adsorption by soils is the increase in concentration of Pb in the equilibrium solution phase as soil Pb concentration increases. The adsorption fits the common Langmiur adsorption isotherm.
Ingested soil enters the stomach and the adsorbed Pb is solubilized by the pH 1 environment. As the digesta enters the small intestine, pH rises to 5-7 within a few cm. Here, soil can play an important role by adsorbing Pb and reducing the solubility of Pb. Thus, because of the adsorption equilibrium process, higher soil Pb concentration should strongly increase the bioavailability of soil Pb. Feeding studies with rats have confirmed this model.
Besides soil Pb concentration, another process may be important in Pb bioavailability. Feeding studies with human volunteers have shown that Pb ingested by fasting humans has 50-80% absorption. When the Pb is ingested with meals, only 2-10% of the Pb is retained. This "meal effect" reduces Pb absorption when food in consumed from about 2 hours before the Pb dose to 3 hours after the dose; this time corresponds to gastric emptying time. Increasing dietary Ca and P reduced retention of Pb by humans, possibly due to Pb coprecipitation with Ca phosphates formed in the duodenum as the digesta pH rises. Although Fe status strongly affects Pb absorption by many animal species, human studies indicated Fe had little affect. However, recent examination of longitudinal childhood blood Pb studies show that blood Pb rises if serum Fe-binding capacity increases (indicative of low body Fe supply in children). Taken together, these findings indicate that nutritional status of Fe, and the Ca and P levels in meals affect absorption of Pb in that meal. Pb ingested in water or paint chips between meals is much more bioavailable than when ingested with meals. However, soil Pb ingested between meals should not have the such greatly increased bioavailability because, in contrast to water or paint chips, soil can adsorb Pb.
Larger particles of metallic Pb and paint Pb have been shown to have lower bioavailability, partly because larger particles are not solubilized as readily as small particles during the short incubation in the stomach. Although powdered PbS is essentially as available as Pb acetate to humans, some forms of mine wastes may be less bioavailable because of the larger particle size. Research is needed to clarify the role of individual soil properties (CaCO3, organic matter, and hydrous Fe oxide concentration) on the bioavailability of soil Pb, and to determine if long term equilibration of Pb in soil allows formation of more stable chemical species of lower bioavailability. It is possible that soils containing CaCO3 could reduce Pb bioavailability because of the higher Ca level, but most soils have substantial adsorbed Ca.
Thus the effect of soil adsorbing Pb in the intestine must be considered in setting limits for maximum soil Pb in areas where children live or play in order to protect children. Because the worst case child may have pica for soil, the ultimate soil Pb limit will depend more on the bioavailability of soil Pb than on the amount of soil ingested.
INTRODUCTION Over the last decade, we have conducted, reported, and reviewed research on potential risks of excessive soil Pb. Our interest began with the increase in urban gardening and questions about the safety of consuming vegetables grown in urban gardens (Chaney et al., 1984). In that work we examined the overall pattern of soil Pb concentration in vegetable gardens in the metropolitan Baltimore, Maryland area (Mielke et al., 1983). Soils in urban areas have become extensively polluted by Pb from automotive emissions, flaking paint and industrial emissions.
In 1986, Chaney and Mielke reviewed the sources of Pb and extent of Pb contamination of urban soils, and analyzed the available data upon which maximum soil Pb limits for habitation by children could be based. These previous papers are detailed reviews of complicated environmental contamination and risk analysis regarding soil Pb.
The present paper considers several aspects of soil Pb risk in further detail. "What is the evidence for chemical species of Pb in Pb-enriched soils and dusts, and do Pb species affect risk?" "What is the mobility of Pb in soil?" And, most important, "How bioavailable is soil Pb to children who ingest soil inadvertently or by pica?"
Chemical speciation usually controls food chain transport and risk from heavy metals in the environment (Chaney, 1988). In simple water Pb toxicity to aquatic organisms, the chemical activity of Pb2+ ions appear to control toxicity (e.g. Freedman et al., 1980). In the case of Pb-rich urban soils and dusts, chemical speciation, particle size, and human nutritional factors appear to each be very important in Pb risk to children. Further, the ultimate question of speciation is the chemical form of Pb in the duodenum when soil is included in the diet with meals or between meals.
Pb Contamination Remains Near the Soil Surface
The major source of Pb in the environment is from its use as an antiknock additive to petrol (Nriagu and Pacyna, 1988). Research has repeatedly shown that small Pb-rich particles reaching the surface of a soil profile largely remain on or near the surface for a prolonged period (Chaney and Mielke, 1986). Good evidence of this is shown in Jordan and Hogan (1975), Preer et al. (1984), Getz et al. (1977), and many other publications. Humans often mix soils to aid growth of plants or lawns, and the Pb is mixed to tillage depth. Individuals often cover old contaminated soil with new soil, or dilute it with soil amendments, sod, or potting soil. Eventually, soil movement by earthworms, or through macropores, etc., allow mixing of surface soil and Pb to deeper strata in the horizon. Human activities are the most important factor affecting the Pb concentration at different soil depths.
Table 1. Effect of depth in soil and distance from a painted
wall on Pb concentration in "houseside" soil in Bowie, Maryland.
Three sides of the house (over 100 years old) were sampled.
Each sample represents a composite of 10 subsamples; samples
dried, sieved < 2 mm, and extracted with 1.0 N HNO3 according
to Chaney et al., (1984).
Side of House
Distance Depth A B C
m cm ------mg Pb/kg dry soil--------
0-1 0 - 5 1050 44700 7330
5 - 10 1060 20600 4680
10 - 15 940 7270 3300
5 0 - 5 431 110 298
5 - 10 404 2020 366
10 - 15 400 2110 286
10 0 - 5 194 1940 730
5 - 10 162 374 686
10 - 15 248 2175 452
A student (Ms. J. Gillespie) working with us characterized the effect of distance from a house, and depth in the profile, on soil Pb concentration. Table 1 shows the results. This location had extreme surface soil Pb enrichment, up to 4.5% Pb, mostly due to paint Pb released from the painted wooden structure.
Perhaps the most important aspect of mobility of soil Pb is its ability to be transported into homes and become part of housedust. Chaney and Mielke (1986) reviewed the available data. It is clear from the strong correlation of soil Pb concentration and housedust Pb concentration in several large surveys that soil Pb contributes to housedust Pb (Thornton et al., 1985;
Jordan and Hogan, 1975). While higher concentrations of Pb are often found for housedust than in soil, the exterior reservoir of Pb is much larger. Interior dust Pb covers surfaces while exterior soil Pb accumulates to considerable depth as noted in Table 1.
After many years research on transport of automotive emission Pb to children, it became clear that much of the transport occurred through ingestion of contaminated housedust rather than inhalation of contaminated air. Respirable size particles quickly agglomerate into particles too large to reach absorption sites in the lung. As with soil Pb, hand to mouth transfer of housedust allows ingestion of persistent environmental Pb contamination.
Pb Concentration in Particle Size Fractions of Soil.
Soil usually has a larger average particle size than housedust or hand dust. The concentration of Pb in soil and dust increases as particle size decreases (Duggan and Inskip, 1985; Duggan et al., 1985; Fergusson and Ryan, 1984; Que Hee et al., 1985; Spittler and Feder, 1979). This is especially important because research has shown that smaller dust particles are more easily transferred to the hands and tend to remain on hands longer. Generally, particles remaining on hands are <50-100 microns in diameter. Thus the higher concentration of Pb in smaller soil/dust particles increases the potential risk of soil Pb (Duggan et al., 1985). As will be noted below, both smaller particle size and greater Pb concentration in the smaller particles both increase the bioavailability of Pb in dust and soil, increasing the risk in yet another way.
Chemical Species of Pb in Pb-Rich Soils. Researchers have used several approaches to identify the chemical species of Pb in soils. This goal was attempted because from many research areas indicated that availability depended on the chemical activity of Pb2+ which in turn depended on the chemical species of Pb present in an environmental sample. The form of Pb emitted by automobiles is the water soluble PbBrCl. Researchers found that this form of Pb is rapidly transformed to PbSO4 in the atmosphere (Biggins and Harrison, 1979).
Four general approaches were taken to characterize the chemical species of Pb in soils, dusts, and street dusts: 1) identify crystalline Pb compounds using X-ray diffraction; 2) determine which compound controls the solubility of Pb in soil solution; 3) extract soils with different reagents and compare results with known compounds added to soils; and 4) add compounds to soil and see if they persist or are transformed to other chemical species.
Because Pb is held near the surface of soil, researchers looked for crystalline Pb compounds in the soil. Olsen and Skogerboe (1975) used density and magnetic separation methods to concentrate the Pb bearing particles of roadside soils. They found crystalline PbSO4 and PbCO3in a calcareous soil, but these accounted for < 5% of the soil Pb. This approach was used to characterize Pb in street dusts by Harrison et al. (1981). They found PbSO4, Pb, Pb3O4, PbOPbSO4, and 2PbCO3Pb(OH)2 in non-magnetic, high density, fine fraction of several street dusts. However, only 0.5 to 2.0% of the Pb existed as a crystalline form susceptible to X-ray diffraction analysis.
Harrison et al. (1981) added 3000-4000 mg Pb/kg of PbSO4 to an uncontaminated soil and incubated the mixture for a short period. They used sequential extractions with several reagents to characterize species of Pb present, and found the extracts were very similar to normal Pb-rich soils rather than to PbSO4 mixed with dry soil. These findings suggest that PbSO4 is quickly converted by soil chemical processes to other, noncrystalline, species of Pb.
The equilibrium solubility approach has been used by several research groups. Santillan-Medrano and Jurinak (1975) first tried this approach and concluded that Pb(OH)2, Pb3(PO4)2, Pb4O(PO4)2, Pb5(PO4)3OH, and PbCO3 should control solubility depending on soil pH and phosphate status. Lindsay (1979) summarized the solubility of Pb in soils compared to theoretical equilibrium solubility of many Pb compounds as a function of soil pH, sulfide, chloride, phosphate, redox etc. Pb4+ cannot persist in soils because it is reduced to Pb2+. Pb cannot persist (e.g. from Pb shot mixed in soil). PbSO4 is too soluble to persist in soils, while PbCO3 was more soluble than Pb5(PO4)3Cl. Nriagu (1974) initially studied the solubility of Pb phosphates, and found evidence to indicate that Pb5(PO4)3Cl should form in soils if adequate phosphate is present, and is the least soluble compound identified. However, when Garcia-Miragaya (1984) examined the solubility of Pb in "old" Pb-rich road-side soils, Pb was generally less soluble than could be accounted for by these known crystalline compounds at equilibrium, but more soluble than Pb5(PO4)3Cl. We believe there are two possible explanations for the inability to demonstrate that Pb solubility is controlled by crystalline compounds: 1) Pb is present in the form of amorphous compounds after only a few years to reach equilibrium; or 2) Pb is adsorbed to strong adsorption sites in the soil and not present in inorganic compounds.
Adsorption of Pb by Soils.
Numerous studies have shown that large amounts of Pb can be adsorbed by soils. This adsorption fits the Langmuir adsorption isotherm, in that as the amount of soil-adsorbed Pb increased, the Pb concentration in the equilibrium solution phase increased. Two groups studies many soils and used multiple linear regression to relate soil properties to the ability of soils to adsorb Pb (Hassett, 1974; Zimdahl and Skogerboe, 1977). The latter authors simplified their multiple regression equation to pH and CEC (which replaced clay and organic matter identified by stepwise regression). Table 2 shows the results of these regressions, with a model calculation for a "typical" pH 7 soil containing 20 milli-equivalents of cation exchange capacity (CEC). CEC is usually correlated with soil clay and organic matter levels, and soil clays are usually covered with a layer of hydrous Fe oxide with strong Pb adsorption ability.
Other researchers have examined adsorption or specific binding of Pb by soil components such as organic matter, hydrous iron oxides, and even manganese oxides (Wolf et al., 1977; McKenzie, 1978; Kinniburgh et al., 1976). In general, humic acids in soil can bind large amounts of Pb very strongly, and the bound Pb is less susceptible to release by acidification than is Pb bound to mineral portions of the soil. However, at the pH of the duodenum, pH 6-7, soil organic matter and Fe oxides can bind large quantities of Pb.
Table 2. Adsorption of Pb by soils fits the Langmuir
adsorption isotherm. Best fit multiple linear regression
models calculated for "typical" soil of cation exchange
capacity (CEC) = 20 milli-equivalents/100 g and pH = 7.
Further, Pb absorption maxima from leaching were 1.6 times
greater than from batch studies.
Recent work by Harter (1979; 1982) has provided further data on soil properties related to adsorption of Pb. Unfortunately, nearly all these studies suffered from an inherent error in the method used to measure adsorption. Most researchers fail to correct soil pH of the soil + metal solution slurry for acidification which results from the metal ion displacing protons bound by the soil. When soils bind 100 μmoles of Pb/g soil, about 100 μmoles of protons are displaced, driving the soil pH down several units. Because significantly more Pb is bound at higher pH than lower pH, the inherent lowering of soil pH during the adsorption capacity test causes a false low estimate of adsorption capacity. One can correct the pH by titrating the mixture with KOH, and then one finds many fold higher Pb adsorption at pH 6-7 than found when a soil initially pH 6-7 is measured for adsorption capacity without pH correction (Harter, 1982). Thus, the very high values noted in Table 2 significantly underestimate potential Pb adsorption capacity. Worse, the role of soil properties (organic matter versus Fe-oxides) may be falsely evaluated because the equilibration took place at pH 4 rather than at pH 6-7. In general, the evidence is very strong that adsorption of Pb by soil accounts for the solubility of Pb in soil, the rapid solubilization of soil Pb upon acidification, and the lack of crystalline Pb compounds in soils.
Adsorption is important in soil in the field in terms of leaching of Pb down the soil profile. However, this process is also important in the small intestines of children who ingest Pb-rich soil. Although soil-bound Pb is solubilized in the pH 1 monogastric stomach (Day et al., 1979), in need not remain soluble during transit of the digesta through the intestines. As the pH rises in the duodenum, soil materials should re-adsorb Pb released by acidification in the stomach. This process should still follow the Langmiur adsorption isotherm, and the relative bioavailability of soil Pb should increase with increasing soil Pb concentration according to the adsorption model.
What Factors Should Influence Risk of Soil Pb? Pb concentration in soil, size of soil particles rich in Pb, chemical species of Pb in soil, and nutritional factors together interact with human behavioural factors in controlling risk from soil Pb. We have reviewed the behavioural factors (Chaney and Mielke, 1986), and must note that children vary remarkably in blood Pb (Pb-B) when exposed to similar Pb sources. Parental supervision, personal habits (mouthing of fingers, hands, toys; chewing fingernails; washing hands), pica behaviour, and quality of nutrition vary among children so greatly that some children can have little risk when they live or play in areas with high soil Pb. However, we must consider the child described by Duggan and Inskip (1985), the average child playing in a normal dirty way. Further, we support the view of the EPA that a proper regulatory evaluation must consider the "most-exposed, most-susceptible individual." In the case of soil Pb risk analysis, this individual is a poorly-supervised child who regularly plays in Pb-rich soil, has pica for soil, and has poor nutrition. This child would ingest much soil Pb, and absorb a higher percentage of this Pb than the well cared for child. Society must protect this child from excessive soil Pb in the urban environment.
Bioavailability of ingested Pb When Pb in soil or dust is ingested by humans, the potential for adverse effects depends on absorption of the ingested soil/dust borne Pb. Research on laboratory animals over many years has characterized the effect of nutritional factors on Pb absorption (Mahaffey, 1982, 1985; Mahaffey and Michaelson, 1980). More recently, adult human Pb isotope absorption studies and Pb balance studies in infants have clarified understanding of human Pb absorption. In addition, several feeding studies using livestock and laboratory animals have directly tested Pb absorption from dietary soil/dust.
Effect of Pb compound and particle size on Pb absorption
Research was conducted to evaluate the bioavailability of Pb in different compounds. Allcroft (1950) reported feeding several compounds to cattle, and found effects of both Pb compound and particle size. In particular, large particle PbS had lower toxicity and caused lower tissue levels of Pb than small particle PbS or other compounds. Although these were long term feeding studies, cattle were studied. Ruminant animals use the anaerobic rumen to process forages, and this anaerobic environment is usually rich in sulfide. Further, the pH of the rumen is near neutral. PbS would not be expected to be dissolved under these conditions. The pH of the stomach fluid in cattle is much higher than that of monogastric organisms such as humans. Gastric fluid is normally pH 3.5-4.5 in cattle, while gastric fluid of humans is usually pH 1. Barltrop and Meek (1975, 1979) studied bioavailability of different Pb compounds and paint pigments (of varied particle size) to rats. Their design used a 48 hour feeding period with very high Pb concentrations. Although their results should provide a good indication of the bioavailability of these compounds in an "acute" protocol, their method should not allow proper evaluation of the effect of dietary constituents on the "chronic" bioavailability of the different Pb compounds. In their work, larger particles had lower bioavailability than smaller particles of several materials. This would appear to result from poorer dissolution of larger particles during the short period of acidic treatment in the stomach, and would be relevant to materials with lower solubility. Compounds which are readily dissolved in weak acid were highly bioavailable. Interestingly, human feeding studies of Rabinowitz et al. (1980) tested the absorption of finely divided PbS to fasting human volunteers. In this test, PbS was highly bioavailable. More recently, Healy et al. (1982) tested the solubility of different particle size preparations of PbS in gastric fluid. They found the expected more rapid solubility of smaller particles. Their work was focused on bioavailability of PbS from cosmetics (e.g. surma) which appear to be transferred to the mouth after hand contact (Healy et al., 1982; Healy, 1984). The implication of these findings for soil Pb are that larger particle size PbS (e.g. galena ore particles dispersed by mining, transport, or smelting) would be expected to have lower bioavailability than other soil Pb. However, grinding of these particles by shoes in the home could produce a fine particle PbS in housedust which would have high bioavailability.