Heavy metal distribution, bioaccessibility, and phytoavailability in long-term contaminated soils from Lake Macquarie, Australia.
Environmental degradation by anthropogenic activities has led to significant build-up of contaminants in the environment, posing risk to ecosystem and human health (Naidu et al. 2003a, 2003b). Consequently, numerous programs have been implemented in most countries to assess the extent of contamination and to initiate appropriate remediation strategies to manage such contamination (Cartwright et al. 1977; Pichtel et al. 1997; Ashley and Lottermoser 1999; Clark et al. 2001).
The North Lake Macquarie area of NSW, Australia, has a long history of environmental impact from industrial activities. Historically, this area has been significantly affected by the Pasminco Metals Smelter, which was in operation from 1897 to 1922 and recommenced operation in 1962. Since that time smelter activities have continued to as recently as September 2003 (Morrison 2003). It was claimed that the smelter was emitting Pb to the air resulting in elevated concentrations in the surrounding soils primarily from aerial deposition (Robinson 1993). High soil Pb levels have long been demonstrated, in numerous countries including Australia, to adversely impact public health. An investigation of the influence of Pb contamination on human health in this area in 2000 revealed that >30% of the children <4 years old had blood Pb levels >10 g Pb/dL (SMH 2000). These results exceed the Australian National Health and Medical Research Council (ANHMRC) Health Investigation level (10 g Pb/dL). A recent investigation demonstrated that blood Pb concentration still exceeded 10gPb/dL in 7% of the children (<5 years old) tested (New South Wales (Hunter New England Area Health Services) 2006). This suggests that there was still a large effect of Pb on the infant population despite an annual decrease in the percentage of infants exceeding the ANHMRC threshold blood Pb concentration since 1991 when the Pb management program was launched.
Literature shows that Pb management strategies in this region were based on total metal concentrations. However, it is now well recognised that metal toxicity and availability is closely related to the fraction of metals present in bioavailable form (Ruby et al. 1993). Bioavailability is often used as the key indicator of potential risk that chemicals pose to environmental and human health. Kelly et al. (2002) defined bioavailability as the fraction of a contaminant in a particular environmental matrix that is absorbed by an organism via a specific exposure route. Thus, an effective Pb management strategy must be based on metal bioavailability rather than total Pb content as has been done in the North Lake Macquarie area. However, considerable controversy exists regarding the index of bioavailability. For instance, heavy metal bioavailability has been linked to solution phase metal concentration (Krishnamurti and Naidu 2000, 2002, 2003), metals extracted using salt solutions (Krishnamurti et al. 2000; Krishnamurti and Naidu 2002; Wenzel et al. 2003), and the partitioning coefficient (Kd), which is a measure of the sorption capacity for a given soil. Several researchers have developed in vitro tests to estimate the bioavailability of a chemical from a soil under simulated gastrointestinal conditions. One method which has gained acceptance among researchers is the simplified physiological-based extraction test (SBET) (Kelly et al. 2002) which is highly correlated with in vivo tests (animal feeding studies) for the assessment of bioavailability of Pb and As (Imber 1993; Ruby et al. 1993; Schoof et al. 1995). Generally, the SBET-extractable heavy metal is called the bioaccessible pool.
During the SBET, only gastric phase absorption is considered for bioaccessibility determination because there was no improvement in estimates of bioaccessibility where obtained when using the additional intestinal phase extraction of the full PBET (physiological-based extraction test) (Kelly et al. 2002).
The objectives of this study were to investigate (a) the distribution of metals in the area impacted by smelter emissions and (b) metal availability. The rationale for using long-term contaminated soils is based on the fact that most studies on metal availability have been conducted using soils freshly spiked with chemicals and such soils have limited significance to long-term contaminated soils (Naidu et al. 1994; Sauve et al. 1997; Krishnamurti and Naidu 2003).
Methods and materials Sampling locations
The region surrounding the Pasminco Metals Smelter has been significantly affected by Pb contamination from smelter operation for many years (Morrison 2003). Air and water degradation due to the reported high levels of Pb present in the soil and housing of the North Lake Macquarie suburbs of Boolaroo, Argenton, and Speers Point was attributed to being a direct consequence of smelter operation (Morrison 2003). Sampling sites were biased towards the area of high Pb levels as they were selected on the basis of previous reports of elevated Pb concentrations at these sites and proximity to residential locations. Samples of surface soil (0-100mm) were taken from an area within a 4-km radius of the smelter including residential areas, public parks, playgrounds, and recreational reserves (Fig. 1). Eight sites, which were expected to be highly contaminated, were selected, together with one control site, which was located almost 10 km from the smelter and therefore thought not to be affected by smelter operations.
Sampling and sample processing
Between 3 and 5 discrete soil samples (0-100mm) were collected from each of the 9 selected sites. The number of samples from each site was determined based on the circumstances existing at that site, including the total area, the terrain, site history, and current usage. The total number of soils sampled was 32 from 9 selected sites, including one sample of smelter slag that was commonly dumped around Lake Macquarie in the early years of smelter operation. Surface soils were collected using a stainless steel soil auger (9 cm in diameter). A total of ~2 kg of bulked soil from 3 or 4 discrete core samples 1 m apart was collected in 'ziplock' plastic bags. Grass, leaves, and roots present in some soil samples were discarded after gentle shaking to remove the soil attached around roots. All soil samples were air-dried and sieved to <2 mm using a stainless steel sieve and stored in plastic bags until analysed.
Routine laboratory analyses
Soil pH and EC were determined on a 1:5 soil:water (w/v) suspension using a combination pH-EC meter (smartCHEM-LABORATORY, TPS) following 1 h equilibration. Soil organic matter content was determined using the Walkely-Black method (Nelson and Sommers 1996) and soil texture was determined using a micro-pipette method (Miller and Miller 1987). Major exchangeable cations were determined by ICP-MS following 1 M N[H.sub.4]OAc extraction. For this, soil (5 g, <2 mm) was extracted with 1 M N[H.sub.4]OAc (50 mL, pH 7) (Sumner and Miller 1996). Total dissolved organic carbon (DOC) was analysed using an automatic total organic carbon analyser (Model 1010, O.I. Analytical) after extracting soil solution. Soil solution was collected following centrifugation. Soil (20g) was weighed into a 20-mL disposable plastic syringe having the tip plugged with acid-washed glass wool. The syringe was inserted in a 50-mL centrifuge tube separated from the base by a PVC spacer. Water was added to the soil to obtain 70% of the maximum water holding capacity (WHC) and the soil was allowed to equilibrate overnight. The resulting soil paste was centrifuged at 905 rcf (2500rpm) for 25 min, and the dissolved organic carbon (DOC) in the extracted solution was measured after filtering the isolated soil solution through 0.45-Bm cellulose acetate disposable filters (Millex[TM], Millipore).
Heavy metal content
Total metal content
Total heavy metal contents were determined by ICP-MS following microwave assisted aqua regia digestion. Air-dried soil (0.5 g, <2 mm) was weighed directly into a teflon digestion vessel and aqua regia (5 mL) was added. The soil suspension was digested in a micro-wave digestion oven (MARS5, CEM) in accordance with Method 3051H (USEPA 1997). After digestion, the soil suspension was decanted into a 50-mL volumetric flask and made up to volume using Milli-Q water and filtering through 0.45-[micro]m filters before analysis. All digestions were conducted in triplicate and each batch included a standard reference material (Montana Soil SRM2711, Certificated by National Institute of Standards & Technology) and blank to validate the digestion operation. Likewise, aqua regia assisted digestion was repeated with soil (<250 [micro]m) for use in determining metal bioaccessibility in conjunction with the SBET extraction results (Kelly et al. 2002). The concentration of heavy metals determined by aqua regia digestion will hereafter be referred to as the total heavy metal concentration.
[FIGURE 1 OMITTED]
Heavy metal bioavailability
Bioaccessible and phytoavailable heavy metal concentrations were estimated using 2 common extraction techniques, SBET for bioaccessibility and 1 M N[H.sub.4]N[O.sub.3] for phytoavailability.
Lead bioaccessibility was estimated using the SBET extraction procedure (Kelly et al. 2002). Soil (1 g, <250 [micro]m) was extracted with a simulated gastric solution (100mL, 0.4 M glycine adjusted to pH 1.50 [+ or -] 0.05 with HCl) which had been equilibrated at 37 [+ or -] 2[degrees]C. The mixture was extracted for 1 h with constant mixing on a rotary suspension shaker at 37[degrees]C. After extraction, the pH was determined and the suspension was subsequently filtered through 0.45-[micro]m filters and stored at 4[degrees]C until analysis. The final pH remained within [+ or -] 0.5 pH units of the initial pH of 1.5.
The phytoavailable metal pool was determined using 1 M N[H.sub.4]N[O.sub.3]. Soil (10 g, <2 mm) was shaken overnight in end-over-end suspension shaker with 1 M N[H.sub.4]N[O.sub.3] (25 mL) in a 50-mL centrifuge tube. The soil suspension was filtered through 0.45-[micro]m disposable filters and stored at 4[degrees]C until analysed.
All the analyses were conducted in duplicate except aqua regia digestion, which was conducted in triplicate and the mean values were collated for use in all data processing. Heavy metals and major cations in extracted solution were determined using ICP-MS.
All the data were statistically analysed using SPSS (SPSS 12.0.1, SPSS Inc.) to examine correlations between metal concentrations extracted using different techniques and for studying the influence of soil properties on metal availability. Total heavy metals, bioaccessible, and 1 M N[H.sub.4]N[O.sub.3] extractable metal concentrations were processed after being log-transformed in order to normalise the distribution of data.
Results and discussion
Physicochemical properties of the soils are summarised in Table 1. Soil [pH.sub.water] ranged from weakly acid (5.1) to weakly alkaline (8.2) with a median value of 6.7. Total clay content ranged from 1.4% to 24.6%. Organic matter (OM) content ranged from 1.1 to 13.8% with a median value of 5%, while DOC in the soil solution ranged from 10 to 407mg/L with a median value of 130mg/L. Plant cover of soil seems to play a strong role in elevating both OM and DOC, as the highest concentrations of OM and DOC were observed in soils from sites LMa03 and LMa08, which were well-developed grassy areas. In comparison, lower concentrations of OM and DOC were observed in the reserve areas (LMa02) which had little plant cover. It is well known that plant roots exude organic carbon as various compounds which are synthesised during photosynthesis and translocated into the roots (Jones 1998). This can result in elevated DOC concentration near roots (Nigam et al. 2001). As expected, the slag sample from LMa02 had the highest pH and the lowest clay, OM, and DOC content. In addition, heavy metal concentrations in this sample exceeded that recorded for other soils (discussed below). Hence, the data from this one sample, essentially a pure slag, were processed separately and has not been included in any regression analysis.
Heavy metal distribution
Similar to previous observations (Galvin et al. 1992), Pb was found to be the main contaminant in this region, as most of the soils studied exceed the Australian Health Investigation Level (HIL) of 300mg/kg for a 'Standard' residential area (Table 2), indicating significant contamination of the soils.
However, the soil Pb concentrations recorded in this study were significantly higher than those previously reported. The elevated Pb concentrations may be attributed to additional deposition from continuous smelting operations since the study conducted by Galvin et al. (1992). They reported the highest value for soil Pb in Boolaroo, adjacent to the smelter, with the concentration decreasing with increasing distance from the smelter (Eqn 1):
log (soil Pb) = -0.002 distance + 9.023 ([r.sup.2] = 0.681, P<0.001)
where distance was in metres and the concentration was in mg/kg.
Similar patterns of Pb distribution were observed in the present study, with the Boolaroo samples, LMa01 and LMa02, showing the highest Pb concentrations among the samples. However, in contrast to Galvin et al. (1992), our study does not show a logarithmic relationship between Pb content and distance from the smelter. This is not surprising given that our sample collection was biased towards areas expected to have elevated Pb concentrations.
Large differences in Pb concentration were observed from discrete samples collected within each site. For example, soil samples from LMa01 were collected from within 10 m of each other, but the Pb concentration differed 10-fold. This is highly indicative of heterogeneity of the sites, probably due to disturbance of the surface soil caused by construction activities in residential areas, such as road work or the widespread use of slag as a material for landfill.
Similar to Pb, both Cu and Zn generally exceeded the environmental investigation level and showed a positive correlation with Pb. This suggests that Cu and Zn were also loaded onto the soil from the smelter. The concentrations of Cu and Zn also exceed the relevant HILs, 1000 and 7000 mg/kg, respectively, for the 'Standard' residential area at some sites. The slag sample (LMa02-1) showed the highest concentrations of Cu, Zn, and Pb.
As expected, the solid-liquid partition coefficient ([K.sub.d]), defined as the ratio of metal in a soil to the metal in the solution (Naidu et al. 1994; Sauve et al. 2000a, 2000b; Krishnamurti and Naidu 2003):
[K.sub.d] = metal concentration in soil/metal concentration in solution
where the metal concentration in the soil is in mg/kg and the dissolved metal concentration is in mg/L, increased with increasing metal content. This may have implications for both metal bioaccessibility and phytoavailability, discussed in the following sections.
Bioaccessibility of Pb
Table 3 shows the variation in SBET (hereafter defined as bioaccessible Pb) and total Pb concentration in the <250 [micro]m soil particles. Bioaccessible Pb ranged from 32 to 100% of the total Pb in soils with low and high metal content soils, respectively. In general, bioaccessible Pb increased with increasing total Pb (Fig. 2). More than 80% of total Pb is bioaccessible in the highly contaminated soils (LMa01 and LMa02). The high bioaccessibility of Pb suggests that contamination in this region poses significant risk to human health. This is also evident from the blood Pb results reported for children (<4 years old) in a survey showing that 30% of those assessed had blood lead levels >10g Pb/DdL (SMH 2000).
[FIGURE 2 OMITTED]
Correlation studies show a weak negative relationship between soil clay content and bioaccessible Pb ([r.sup.2] = 0.12, P<0.1), suggesting that increasing clay content has the potential to reduce Pb bioaccessibility in these soils. However, there was no correlation between soil OM and bioaccessible Pb. This may be attributed to the large Pb content of soils and subsequent super-saturation of OM binding sites.
There was a significant relationship ([r.sup.2]=0.60, P<0.001) between bioaccessible Pb as estimated by SBET and 1 M N[H.sub.4]N[O.sub.3] extractable Pb defined as phytoavailable Pb. The slope of the regression was 0.99 and therefore within experimental error of unity as might be expected from a 1 : 1 relationship between bioaccessible and 1 M N[H.sub.4]N[O.sub.3] extractable Pb. This suggests that 1 M N[H.sub.4]N[O.sub.3] extractable Pb could be a reliable measure of bioaccessibility, although further detailed studies using in vivo techniques are needed to substantiate these conclusions.
Phytoavailable metal pools
Widely different extractants have been used to estimate metal phytoavailability in soils. To date there is no universally accepted extraction procedure for assessing phytoavailable metals. For instance, ammonium chloride (1 M N[H.sub.4]Cl) is the common extractant in Canada (Krishnamurti et al. 1995), ammonium nitrate (1 M N[H.sub.4]N[O.sub.3]) in Europe (DIN 1995), and sodium chloride (1 M NaCl) in Switzerland. Given the strong positive correlations between extractable metals and plant metal contents, metals extracted using these methods have often been defined as the phytoavailable fraction. Even though metals extracted with 1 M N[H.sub.4]Cl generally show significant correlation with plant metal uptake (Krishnamurti et al. 2000), extraction with 1 M N[H.sub.4]N[O.sub.3] was preferred given that Cd was one of the metals of interest and [Cl.sup.-] forms stable Cd-chloro complexes (Candelaria and Chang 1997). Ligand complexation is typically a stronger reaction than cation exchange and consequently greater complexation promotes mobility. Along with these extractants, numerous studies (Naidu 2004) demonstrate strongly significant relationship between solid-liquid partition coefficient and plant metal contents. In the following sections we investigate the relationship between soil properties and phytoavailable metal pools as assessed by 1 M N[H.sub.4]N[O.sub.3].
Correlation studies showed strongly significant relationship between phytoavailable Cd (1 M N[H.sub.4]N[O.sub.3] extractable Cd) and total Cd content ([r.sup.2]=0.83, P<0.001) (Fig. 3 and Table 4). The phytoavailable Cd ([r.sup.2]=0.47, P<0.001) increased with decreasing [K.sub.d] (Fig. 4a). This suggests that as the metal binding capacity decreases the amount of Cd available for plant uptake increases (discussed below).
Numerous investigators have demonstrated the role of soil pH on soil Cd sorption (Naidu et al. 1994, 1997; Sauve et al. 1997; Appel and Ma 2002; Collins et al. 2003). Soil pH affects metal sorption/desorption by changing the surface chemical properties of soil and metal speciation directly (Naidu et al. 1994; Gray et al. 1998, 1999). In general, Cd desorption decreases as soil pH increases, which is supported by the observed strong positive correlation between soil pH and [K.sub.d] (Fig. 4b). This may be attributed to the enhanced binding of Cd due to the increase in the soil surface negative charge density with increasing pH.
[FIGURE 3 OMITTED]
Organic matter and DOC can contribute significantly to metal phytoavailability in solution through metal-organic complex interactions (Fotovat and Naidu 1998; Naidu and Harter 1998; Harter and Naidu 2001). Krishnamurti et al. (2000) observed a weak but significant ([r.sup.2]=0.48, P<0.1) relationship between total Cd (by HF-HCl[O.sub.4] extraction) and organic C in soil. In addition, Cd in saturation paste extracts was positively related with organic C in soil (Krishnamurti and Naidu 2003). However, contrary to these published studies, there was no correlation between phytoavailable Cd and OM or DOC in the present study. This may be attributed to (a) the widely different metal loading of soils under study, (b) differences in Cd binding between freshly spiked soils used by Krishnamurti and Naidu (2003) and relatively long-term contaminated soils, and (c) the presence of other heavy metals that have higher metal--organic stability constants (discussed below). It is also likely that the lack of a relationship between phytoavailable Cd and DOC may be due to the presence of insufficient concentrations of organic ligands in soil solution (McBride et al. 1997).
Multiple regression analyses have long been used to investigate the role of soil properties on soluble heavy metal dynamics in soils (Qian et al. 1996; Sauve et al. 2000a). Using step-up regression, the present study shows that the most important variables to predict phytoavailable Cd pool were total Cd ([r.sup.2]=0.83) and soil pH ([r.sup.2]=0.66) (Table 4). As discussed above, the Ka is also strongly correlated with phytoavailable Cd. Other parameters such as OM and DOC did not improve the regression. Similar findings were reported by Krishnamurti and Naidu (2003) and Sauve et al. (2000a), who found that soil organic matter did not improve the predictability of soil solution Cd concentrations.
Correlation studies show a strong significant relationship between phytoavailable Zn and soil pH ([r.sup.2]=0.84, P < 0.001). However, there was no significant relation between soil OM, DOC, and phytoavailable Zn. Step-down multiple regression analyses showed a significant relationship between phyto-available Zn, soil pH, and total Zn (Table 4).
Sauve et al. (1997) reported a strongly significant correlation ([r.sup.2] = 0.86, P < 0.001) between 0.01 M Ca[Cl.sub.2] extractable Cu and total Cu, concluding that the relationship was an outcome of long-term contamination during World War II. More recently, Kachur et al. (2003) reported that in long-term contaminated soils, natural soil processes allow equilibrium to be established resulting in conversion of free [Cu.sup.2+] into less available forms. In the present study, phytoavailable Cu was positively correlated with total soil Cu ([r.sup.2] = 0.63, P<0.001), although the relation was not as strong as that reported by Kachur et al. (2003). This may be attributed to (a) continuous deposition of new Cu due to smelter activities, and (b) the presence of high amount of other heavy metals such as Cd, Zn, and Pb which may compete with Cu for binding sites.
[FIGURE 4 OMITTED]
Complexation reactions with organic ligands in the soil solution is an important mechanism for controlling Cu solubility, with the increase of Cu in soil solution being attributed to DOC (Naidu et al. 1997; Gao et al. 2003; Qin et al. 2004). However, this is not generally consistent for soils with widely different soil properties. For instance, Sauve et al. (1997) found no significant effects of OM on Cu solubility and this result has been supported by other studies (McBride et al. 1997). In the present study, DOC was only weakly correlated with phytoavailable Cu ([r.sup.2]=0.15, P<0.1). This may be attributed to the presence of other heavy metals Pb and Zn both of which would compete for binding sites present on DOC.
The solubility of soil organic matter and organo-metallic complexes are pH-dependent. In the present study [K.sub.d] for Cu decreased with increasing DOC ([r.sup.2] = 0.50, P < 0.1) at pH values >7, while below pH 7 there was no significant relationship between these parameters. This indicates that Cu solubility could be controlled by DOC >pH 7. However, there was only a weak correlation ([r.sup.2]=0.20, P<0.1) between soil pH and phytoavailable Cu. This is consistent with the study by Sauve et al. (1997) and the known insensitivity of Cu solubility to pH, even though free [Cu.sup.2+] activity has a greater sensitivity to pH (McBride et al. 1997; Sauve et al. 1997). The total Cu concentration in the soil had a significant influence on the phytoavailable Cu (Table 4).
According to Sauve et al. (2000a), large concentrations of Pb in solution would promote adsorption and increase the [K.sub.d], indicating high affinity for Pb to the solid phase. This naturally assumes that the soil is not already saturated with metals and that there is available adsorption sites for Pb. However, contrary to that observation, there was no significant relationship between total Pb in soil and Ka in this study. This may be attributed to saturation of sorption sites on soil particle surface and any excess Pb being present in mineral form. Further investigations are in progress to establish the nature of free Pb mineral and will be published in a separate paper. The phytoavailable Pb, however, increased with a decrease in partition coefficient ([r.sup.2]=0.47, P<0.001) (Fig. 5). This suggests [K.sub.d] could play a significant role in controlling the phytoavailability of Pb in these soils. Lead phytoavailability was strongly correlated with soil pH ([r.sup.2]=0.70, P<0.001) suggesting that soil pH is the main factor governing Pb solubility.
McBride et al. (1997) demonstrated a strong relationship between Pb and key soil properties. They suggested that organic matter was significant in predicting Pb solubility and that Pb mobility could be controlled by DOC in soil solution. A more recent study using metal sorption test in tropical soils concluded that organic matter was not the critical factor in Pb sorption (Appel and Ma 2002). Their study was, however, limited to metal contents that are often less than those found in soils from highly contaminated sites in which metal availability is mainly controlled through precipitation and dissolution reactions rather than complexation or cation exchange at adsorption sites. In laboratory studies, metal interactions are generally controlled by sorption and complexation reactions (McBride et al. 1997).
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
Heave metal interaction in soils
When a mixture of heavy metals is loaded onto a soil, individual heavy metal availability can be influenced by competition between metals for the limited number of available binding sites. Each individual heavy metal has different affinity for the deprotonated negatively charged surface sites of soil components, which generally follows the Irving-Williams order:
[Hg.sup.2+] > [Pb.sup.2+] > [Cu.sup.2+] > [Zn.sup.2+] > [Ni.sup.2+~] [Co.sup.2+] > [Cd.sup.2+]
In many studies, Zn and Cd have shown much higher availability than Pb and Cu. This is due to the higher affinity of Pb and Cu for sorption sites relative to Zn and Cd as shown above in the Irving-Williams series. If several heavy metals coexist in the soil, the soluble phase concentrations are likely to be higher for those metals having low stability constants, such as Zn and Cd, while metals such as Pb and Cu which have relatively high stability constant will contribute less to soluble phase concentrations. In the present study, phytoavailable Cd was positively correlated with phytoavailable Pb ([r.sup.2]=0.89, P<0.001) (Fig. 6a), resulting in a decrease in [K.sub.d] for Cd as the phytoavailable Pb increased ([r.sup.2] = 0.50, P< 0.001) (Fig. 6b).
This is also supported by the marked increase in phytoavailable Cd, which exceeded the solid phase Cd in increasing ratio suggesting that Cd adsorption onto soil components was diminished by the increasing amount of Pb present. In addition, the relationship between [K.sub.d] of Cd and phytoavailable Pb was significantly improved with input of soil pH ([r.sup.2]=0.80, P<0.001) (Fig. 7), implying phytoavailability of Cd was highly influenced by both the presence of Pb and soil pH.
Inclusion of phytoavailable Cu improved the relationship between [K.sub.d]-Cd and phytoavailable Pb ([r.sup.2] = 059, P< 0.001). As shown in Fig. 8, an increase in phytoavailable Pb increases phytoavailable Cu and together this decreased [K.sub.d]-Cd, confirming that heavy metal phytoavailability is influenced by both soil properties such as pH and the presence of other metals. A similar effect of Pb on Zn phytoavailability was recorded. As the phytoavailable Pb concentration increased, the [K.sub.d] Zn decreased ([r.sup.2]=0.29, P<0.1), leading to substantial increase in phytoavailable Zn. This increase in Zn is not surprising given that the Zn binding to soil colloids would be reduced due to competition with Pb. In contrast to Cd and Zn, Cu increased in both the phytoavailable pool and [K.sub.d] with increasing Pb. This may be attributed to the similar binding mechanism and capacity of Cu and Pb for soil colloids.
From these results, it can be concluded that any fresh deposition of Pb onto the soil may subsequently increase the solubility of soil surface bound Zn and Cd by competing with these elements for available binding sites. The effect of such mixed systems including more than one heavy metal must be considered when assessing the risks posed by contaminants at long-term contaminated sites.
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
Soils surrounding the Pasminco Metals Smelter were investigated for metal distribution, bioaccessibility, and phytoavailability together with the key soil properties commonly known to influence metal solubility and mixed metals interaction. The soils studied were expected to have been subjected to continuous heavy metal deposits from the smelter and were therefore considered to be long-term contaminated soils.
The primary contaminant in the area was Pb, with 66% of the samples studied exceeding the HILs value of 300mg/kg. Secondary contaminants were Cu and Zn. Simultaneous contamination of Pb, Cu, and Zn probably originated from continual emission of metals from the smelter over an extended period. The exceedance of the HILs for Pb indicates a potential health concern to humans and biota, which was supported by the high bioaccessibility of Pb determined here by the SBET. Bioaccessible Pb ranged from 32 to 100% and increased with total soil Pb ([r.sup.2]=0.98). This suggests that freshly deposited contaminants were not well sorbed to the soils and that super-saturation of the binding sits of the soil was potentially occurring.
While neither organic matter nor dissolved organic carbon showed any major contribution to heavy metal phytoavailability, soil pH was the key soil property which controlled the phytoavailability of Cd, Zn, and Pb. Phytoavailability of all heavy metals studied decreased with increasing soil pH. Heavy metals that coexist in soil compete for available binding sites on a soil, resulting in an increase in phytoavailability of heavy metals such as Cd which have relatively low stability constants, when they compete with heavy metal such as Pb which have higher stability constants. In the present study, an increase in phytoavailable Pb also resulted in an increased phytoavailability of Cd and this correlation was improved with input from soil pH as a regression parameter.
We thank the Department of Education, Science and Technology (DEST), Australia, for the IPRS PhD scholarship to Kwon Rae, and Cooperative Research Centre for Contamination Assessment and Remediation of Environment (CRC-CARE) for providing significant funds towards the research to the authors.
Manuscript received 14 March 2008, accepted 20 October 2008
Appel C, Ma L (2002) Concentration, pH, and surface charge effects on cadmium and lead sorption in three tropical soils. Journal of Environmental Quality 31(2), 581-589.
Ashley PM, Lottermoser BG (1999) Arsenic contamination at the Mole River mine, northern New South Wales. Australian Journal of Earth Sciences 46, 861-874. doi: 10.1046/j.1440-0952.1999.00748.x
Candelaria LM, Chang AC (1997) Cadmium activities, solution speciation, and solid phase distribution of Cd in cadmium nitrate and sewage sludge-treated soil systems. Soil Science 162, 722-732. doi: 10.1097/ 00010694-199710000-00005
Cartwright R, Merry RH, Tiller KG (1977) Heavy metal contamination of soils around a lead smelter at Port Pirie, South Australia. Australian Journal of Soil Research 15, 69-81. doi: 10.1071/SR9770069
Clark MW, Walsh SR, Smith JV (2001) The distribution of heavy metals in an abandoned mining area; a case study of Strauss Pit, the Drake mining area, Australia: Implications for the environmental management of mine sites. Environmental Geology 40(6), 655-663. doi: 10.1007/s002549900073
Collins RN, Merrington G, McLaughlin MJ, Morel JL (2003) Organic ligand and pH effects on isotopically exchangeable cadmium in polluted soils. Soil Science Society of America Journal 67, 112-121.
DIN (Deutsches Institue fur Normung) (1995) 'Soil quality extraction of tmee elements with ammonium nitrate solution.' DIN 19730 (Beuth Verlag: Berlin)
Fotovat A, Naidu R (1998) Changes in composition of soil aqueous phase influence chemistry of indigenous heavy metals in alkaline soils and acidic soils. Geoderma 84, 213-234. doi: 10.1016/S0016-7061(97) 00130-4
Galvin J, Stephenson J, Waller G, Loughran B, Clarke R, Wlordarczyk W (1992) 'A report on the health risk from lead in the environment in Boolaroo, Argenton and Speers Point.' (The Public Health Unit-Hunter Area Health Service: Lake Macquarie, NSW)
Gao Y, He J, Ling W, Hu H, Liu F (2003) Effects of organic acids on copper and cadmium desorption from contaminated soils. Environment International 29(5), 613-618. doi: 10.1016/S0160-4120(03)00048-5
Gray CW, McLaren RG, Roberts AHC, Condmn LM (1998) Sorption and desorption of cadmium from some New Zealand soils: effect of pH and time. Australian Journal of Soil Research 32, 192-216.
Gray CW, McLaren RG, Roberts AHC, Condron LM (1999) Solubility, sorption, and desorption of added cadmium in relation to properties of soils in New Zealand. European Journal of Soil Science 50, 122-137. doi: 10. 1046/j.1365-2389.1999.00221.x
Harter RD, Naidu R (2001) An assessment of environmental and solution parameter impact on trace-metal sorption by soils. Soil Science Society of America Journal 65(3), 597-612.
Imber BD (1993) 'Development of a physiologically relevant extraction procedure.' (CB Research International Corporation: Sydney)
Jones DL (1998) Organic acids in the rhizosphere--a critical review. Plant and Soil 205, 2544. doi: 10.1023/A:1004356007312
Kachur AN, Arzhanova VS, Yelpatyevsky PV, Von Braun MC, von Lindem IH (2003) Environmental conditions in the Rudnaya River watershed--a compilation of Soviet and post-Soviet era sampling around a lead smelter in the Russian Far East. The Science of the Total Environment 303(1-2), 171-185. doi: 10.1016/ S0048-9697(02)00351-0
Kelly ME, Brauning SE, School RA, Ruby MV (2002) 'Assessing oral bioavailability of metals in soil.' (Battelle Press: Ohio)
Krishnamurti GSR, Huang PM, Van Rees KCJ, Kozak LM, Rostad HPW (1995) A new test method for the determination of plant-available cadmium in soils. Communications in Soil Science and Plant Analysis 26(17&18), 2857-2867. doi: 10.1080/001036295 09369493
Krishnamurti GSR, Naidu R (2000) Speciation and phytoavailability of cadmium in selected surface soils of South Australia. Australian Journal of Soil Research 38, 991-1004. doi: 10.1071/SR99129
Krishnamurti GSR, Naidu R (2002) Solid-solution speciation and phytoavailability of copper and zinc in soils. Environmental Science & Technology 36(12), 2645-2651. doi: 10.1021/es001601t
Krishnamurti GSR, Naidu R (2003) Solid-solution equilibria of cadmium in soils. Geoderma 113(1-2), 17-30. doi: 10.1016/S0016-7061(02) 00313-0
Krishnamurti GSR, Smith LH, Naidu R (2000) Method for assessing plant-available cadmium in soils. Australian Journal of Soil Research 38, 823-836. doi: 10.1071/SR99122
McBride M, Sauve S, Hendershot W (1997) Solubility control of Cu, Zn, Cd and Pb in contaminated soils. European Journal of Soil Science 48, 337 346. doi: 10.1111/j.1365-2389.1997.tb00554.x
Miller WP, Miller DM (1987) A micro pipette method for soil mechanical analysis. Communications in Soil Science and Plant Analysis 18, 1-15. doi: 10.1080/00103628709367799
Morrison AL (2003) An assessment of the effectiveness of lead pollution reduction strategies in North Lake Macquarie, NSW, Australia. The Science of the Total Environment 303(1-2), 125-138. doi: 10.1016/ S0048-9697(02)00353-4
Naidu R (2004) Unified approach to assessing contaminant bioavailability in soils. In 'Proceeding of the 3rd International Workshop on Chemical Bioavailability in the Terrestrial Environment'. Adelaide, S. Aust. pp. 7-8.
Naidu R, Bolan NS, Kookana RS, Tiller KG (1994) Ionic strength and pH effects on surface charge and Cd sorption characteristics of soils. Journal of Soil Science 45, 419-429. doi: 10.1111/j.1365-2389.1994. tb00527.x
Naidu R, Harter RD (1998) Effect of different organic ligands on cadmium sorption by and extractability from soils. Soil Science Society of America Journal 62, 644-650.
Naidu R, Kookana RS, Sumner ME, Harter RD, Tiller KG (1997) Cadmium sorption and transport in variable charge soils: A review. Journal of Environmental Quality 26, 602-617.
Naidu R, Megharaj M, Owens G (2003a) Recyclable urban and industrial waste--benefits and problems in agricultural use. In 'Managing soil quality--challanges in modern agriculture.' (Eds P Schjonning, S Emholt, BT Christensen) (CABI Publishing, CABI International: Wallingford, UK)
Naidu R, Rogers S, Gupta VVSR, Kookana RS, Bolan NS, Adriano DC (2003b) Bioavailability of metals in the soil plant environment and its potential role in risk assessment. In 'Bioavailability, toxicity and risk relationships in ecosystems'. (Eds R Naidu, S Rogers, VVSR Gupta, RS Kookana, NS Bolan, DC Adriano) (Science Publishers Inc.: New Hampshire)
Nelson DW, Sommers LE (1996) Total carbon, organic carbon, and organic matter. In 'Methods of soil analysis: Agronomy'. (Ed. AL Page) pp. 961-1110.
New South Wales (Hunter New England Area Health Services) (2006) North Lake Macquarie blood lead monitoring service. Available at: www1.hnehealth.nsw.gov.au/hneph/Lead/BloodLeadTesting.pdf
Nigam R, Srivastava S, Parakash S, Srivastava MM (2001) Cadmium mobilization and plant availability-the impact of organic acids commonly exuded from roots. Plant and Soil 230, 107-113. doi: 10.1023/A:1004865811529
Pichtel J, Sawyerr HT, Czarnowska K (1997) Spatial and temporal distribution of metals in soils in Warsaw, Poland. Environmental Pollution 98(2), 169-174. doi: 10.1016/S0269-7491(97)00131-0
Qian J, Wang ZJ, Shan XQ, Tu Q, Wen B, Chen B (1996) Evaluation of plant availability of soil trace metals by chemical fractionation and multiple regression analysis. Environmental Pollution 91(3), 309-315. doi: 10.1016/0269-7491(95)00066-6
Qin F, Shan XQ, Wei B (2004) Effects of low-molecular-weight organic acids and residence time on desorption of Cu, Cd, and Pb from soils. Chemosphere 57(4), 253-263. doi: 10.1016/j.chemosphere. 2004.06.010
Robinson D (1993) Contaminated land at the Newcastle Lead Smelter and the role of community groups in integrated contaminated land management. In 'Contaminated sites in Australia'. (Eds G Rowe, S Seidler) pp. 87-94. (Allen and Unwin: St Leonards, NSW)
Ruby MW, Davis A, Link TE, Schoof R, Chaney RL, Freeman GB, Bergstrom P (1993) Development of an in vitro screening test to evaluate the in vivo bioaccessibility of ingested mine-waste lead. EnvironmentalScience & Technology 27(13), 2871-2877. doi: 10.102l/ es00049a030
Sauve S, Hendershot W, Allen HE (2000a) Solid-sohitiun partitioning of metals in contaminated soils: dependence on pH, total metal burden, and organic matter. Environmental Science & Technology 34(7), 1125-113l. doi: 10.1021/es9907764
Sauve S, McBride MB, Norvell WA, Hendershot WH (1997) Copper solubility and speciation of in situ contaminated soils: Effects of copper level, pH and organic matter. Water, Air, and Soil Pollution 100(1-2), 133-149. doi: 10.1023/A:1018312109677
Sauve S, Norvell WA, McBride M, Hendershot W (2000b) Speciation and complexation of cadmium in extracted soil solutions. Environmental Science & Technology 34(2), 291-296. doi: 10.1021/es990202z
Schoof RA, Butcher MK, SeUstone C, Ball RW, Fricke JR, Keller V, Keehn B (1995) An assessment of lead absorption from soil affected by smelter emissions. Environmental Geochemistry and Health 17, 189-199. doi: 10.1007/BF00661331
SMH (2000) Director of the Hunter Public Health Unit. Reported by C Delton in the Sydney Morning Herald; 25th September 2000.
Sumner ME, Miller WP (1996) Cation exchange capacity and exchange coefficients. In 'Methods of soil analysis. Part 3: Chemical method'. (Eds DL Sparks, AL Page, PA Helmke, RH Loeppert, PN Soltanpour, MA Tabatabai, CT Johnston, ME Summer) pp. 1201-1230. (SSSA and ASA: Madison, W1)
USEPA (1997) 'Method 3051a: Microwave assisted acid dissolution of sediments, sludges, soils, and oils.' 2nd edn. (U.S. Environmental Protection Agency, U.S. Government Printing Office: Washington, DC)
Wenzel WW, Unterbrunner R, Sommer P, Sacco P (2003) Chelate-assisted phytoextraction using canola (Brassica napus L.) in outdoors pot and lysimeter experiments. Plant and Soil 249(1), 83-96. doi: 10.1023/ A: 1022516929239
Kwon-Rae Kim (A,B), Gary Owens (A), and Ravi Naidu (A,B,C)
(A) Centre for Environmental Risk Assessment and Remediation, University of South Australia, Mawson Lakes, SA 5095, Australia.
(B) Cooperative Research Centre for Contamination Assessment and Remediation of the Environment, PO Box 486, Salisbury South, SA 5106, Australia.
(C) Corresponding author. Email: Ravi.Naidu@crccare.com
Table 1. Characteristics of the soils studied Sample pH EC Sites No. ([micro]S/cm) LMa01 1 6.7 90 2 7.0 81 3 5.1 77 4 5.8 82 LMa02 1 8.7 40 2 7.3 340 3 7.5 52 LMa03 1 6.5 30 2 6.1 49 3 5.8 41 4 6.0 142 5 7.1 91 LMa04 1 7.4 75 2 6.9 50 3 6.5 60 4 7.1 97 LMa05 1 7.2 131 2 8.2 177 3 7.7 153 LMa06 1 5.6 38 2 5.6 43 3 5.2 36 LMa07 1 6.7 36 2 6.9 46 3 6.6 57 LMa08 1 7.0 88 2 7.8 349 3 6.5 82 4 6.3 50 LMa09 1 6.0 43 2 5.7 40 3 5.8 67 Sample Clay OM DOC Sites (%) (mg/L) LMa01 7.1 12.5 147 3.8 3.8 180 2.9 13.8 83 8.7 4.8 95 LMa02 1.4 1.1 10 11.1 3.1 60 12.8 3.4 29 LMa03 3.0 17.0 118 6.6 6.0 154 6.9 8.0 260 10.6 10.9 210 9.5 8.6 82 LMa04 13.6 4.1 124 14.0 3.6 102 19.9 4.2 149 20.1 4.5 390 LMa05 14.0 5.0 94 10.7 5.2 118 11.2 4.6 95 LMa06 24.6 8.1 291 4.6 7.0 94 19.8 3.7 154 LMa07 5.9 3.7 135 7.2 6.2 110 5.6 7.2 248 LMa08 5.3 7.1 199 6.7 5.5 150 8.4 10.1 287 6.9 8.6 407 LMa09 5.4 3.4 122 5.9 2.5 82 16.1 5.8 181 Sample Exchang. cations (mg/kg) Sites Mg K Ca LMa01 215 313 2500 81 180 1200 122 207 1150 73.9 188 630 LMa02 12.0 4.0 2260 190 151 1580 235 94.7 1293 LMa03 60.7 108 390 122 196 730 103 133 750 282 332 1003 207 203 1520 LMa04 301 159 1309 352 150 677 525 159 690 535 279 1090 LMa05 300 203 1590 307 169 2300 239 184 1820 LMa06 670 400 1510 134 327 411 313 183 332 LMa07 120 180 640 145 162 830 237 171 1010 LMa08 167 304 1650 233 300 3500 298 350 1760 240 225 1600 LMa09 200 183 620 158 122 450 219 250 760 Table 2. Heavy metal concentrations (mg/kg) of soils by aqua regia digestion Samples Cu Zn Cd Pb Sites No. LMa01 1 400 4960 65 3970 2 90 1400 12 630 3 560 5900 86 6300 4 131 1600 16 1170 LMa02 1 11 700 90 000 17 58 000 2 730 15 900 10 1440 3 1500 31 000 14 2800 LMa03 1 70 2400 3.1 350 2 110 2700 8 770 3 101 1600 16 1290 4 76 1200 10.5 1070 5 460 38 000 19 2400 LMa04 1 7.6 81 0.4 44 2 7 42 0.5 41 3 6.4 55 0.5 46 4 7.1 50 0.3 41 LMa05 1 240 4000 1.1 550 2 790 12 000 1.1 1700 3 420 7000 1.6 1000 LMa06 1 44 660 3.5 235 2 43 230 3.1 160 3 11 170 1.7 110 LMa07 1 520 14 500 1.6 660 2 1200 29 000 1.3 3100 3 570 12 000 1.2 1200 LMa08 1 590 21 000 1.3 6400 2 200 8000 1.0 800 3 78 3600 1.7 390 4 10.2 310 1.1 79 LMa09 1 7 5 0.10 8 2 6.1 54 0.06 9 3 11 230 0.21 47 Table 3. Lead concentrations of soils (<250 [micro]m) by aqua regia digestion and SBET Samples Total (A) SBET (B) Fraction (C) Sites No. (mg/kg) (mg/kg) (%) LMa01 1 4610 4300 93 2 1400 1150 82 3 7100 5850 83 4 1293 1160 90 LMa02 1 110 000 107 000 100 2 1340 1220 91 3 2410 1900 80 LMa03 1 620 410 67 2 964 800 83 3 1530 1120 73 4 910 900 l00 5 1507 1070 71 LMa04 1 49 29 59 2 32 22 68 3 45 25 56 4 38 21.3 55 LMa05 1 350 250 70 2 840 630 70 3 520 390 75 LMa06 1 252 117 47 2 150 50 33 3 124 83 67 LMa07 1 350 250 70 2 970 1000 100 3 563 520 92 LMa08 1 4420 3300 75 2 558 350 63 3 498 200 40 4 86 63 73 LMa09 1 16 7 43 2 19 5.9 32 3 59 26 44 (A) Lead concentrations digested with aqua regia with <250 [micro]m soil. (B) Simplified physiological-based extraction test (Kelly et al. 2002). (C) Fraction of Pb by SBET to total Pb contents. Table 4. Linear regression of phytoavailable pool (PP) of Cu, Zn, Cd, and Pb against total (T) concentrations of the metals combined with soil pH, OM, and DOC Constant Parameters [R.sup.2] Log (PP-Cd) = 1.675 + 1.186 Log (T-Cd) 0.832 = 4.551 + 1.144 Log (T-Cd) - 0.435 Soil pH 0.947 = 5.777 + 1.225 Log (T-Cd) - 0.468 Soil pH - 0.659 Log (OM) 0.962 = 5.784 + 1.224 Log (T-Cd) - 0.469 Soil pH - 0.655 Log (OM) + 0.005 Log (DOC) 0.962 Log (PP-Zn) = 2.767 + 0.501 Log (T-Zn) 0.350 = 7.768 + 0.788 Log (T-Zn) - 0.899 Soil pH 0.887 = 7.238 + 0.764 Log (T-Zn) - 0.869 Soil pH +0.262 Log (OM) 0.890 = 7.142 + 0.772 Log (T-Zn) - 0.871 Soil pH +0.218 Log (OM) + 0.073 Log (DOC) 0.891 Log (PP-Cu) = 1.356 + 0.526 Log (T-Cu) 0.625 = 1.982 + 0.569 Log (T-Cu) - 0.108 Soil pH 0.647 = 1.375 + 0.536 Log (T-Cu) - 0.076 Soil pH +0.294 Log (OM) 0.658 = 1.228 + 0.560 Log (T-Cu) - 0.083 Soil pH +0.208 Log (OM) + 0.130 Log (DOC) 0.660 Log (PP-Pb) = 0.204 + 0.882 Log (T-Pb) 0.413 = 5.782 + 1.128 Log (T-Pb) - 0.964 Soil pH 0.833 = 6.949 + 1.217 Log (T-Pb) - 1.009 Soil pH - 0.626 Log (OM) 0.842 = 6.563 + 1.264 Log (T-Pb) - 1.016 Soil pH - 0.849 Log (OM) + 0.306 Log (DOC) 0.845
|Printer friendly Cite/link Email Feedback|
|Author:||Kim, Kwon-Rae; Owens, Gary; Naidu, Ravi|
|Publication:||Australian Journal of Soil Research|
|Date:||Mar 1, 2009|
|Previous Article:||Evidence for biocycling from Ba/Ca, Sr/Ca, and [sup.87]Sr/[sup.86]Sr in soils (Red Brown Earths) from South Australia.|
|Next Article:||Distribution of carbon, nitrogen, phosphorus, and sulfur in the soil in a multiple habitat system in India.|