| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a University of Idaho, Moscow, ID 83844-2339
b Patuxent Wildlife Research Center, Laurel, MD 20708
* Corresponding author (dgstrawn{at}uidaho.edu)
Received for publication August 18, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: GI, gastrointestinal • ICP-AES, inductively coupled plasma–atomic emission spectrometer • PBET, physiologically based extraction test • W-PBET, waterfowl PBET
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Due to the risks of Pb poisoning to wildlife and humans in the Lower Coeur d'Alene River Basin, and the vast land area that needs to be remediated, addition of phosphates to the soils has been proposed as an in situ remediation strategy. Application of phosphate amendments changes the Pb chemistry through formation of sparingly soluble Pb phosphates (Ruby et al., 1994; Yang et al., 2001; Zhang et al., 1997). Several studies have demonstrated decreased (bio)availability of Pb minerals in P-treated soils (Davis et al., 1993; Ruby et al., 1994; Yang et al., 2001; Zhang et al., 1997). Due to the variable environmental conditions and dynamic nature of soil biogeochemistry, however, an assessment of the in situ remediation strategy must be done on a case-by-case basis and must take into account the environmental factors controlling contaminant speciation (e.g., reduction and oxidation cycling).
There are several different approaches for measuring bioavailability. In vivo tests use an animal to measure absolute bioavailability and toxicity. In vitro tests are performed outside the organism. For toxicological studies, the outcome of an in vitro test must be correlated to animal bioavailability and toxicity. One type of in vitro test is a PBET that incorporates gastrointestinal (GI) tract parameters representative of a particular species. Initially, PBET models were designed to simulate the human GI tract, which includes stomach and intestinal phases (Basta and Gradwohl, 2000; Medlin, 1997; Oomen et al., 2002; Rodriguez et al., 1999; Ruby et al., 1996; Ruby et al., 1993; Schroder et al., 2003). Absolute bioavailability is the amount of a substance absorbed into the organism's tissue via a particular route of exposure divided by the total amount administered (USEPA, 1999). In vitro bioavailability (bioaccessibility) is defined as the solubility of soil metal in simulated stomach and intestinal solutions divided by total metal in the soil (Berti and Cunningham, 1997). Physiologically based extraction tests differ from other soil extraction tests (e.g., Toxicity Characteristic Leaching Procedure (TCLP), Mehlich plant nutrient availability test, etc.) because they incorporate physiological parameters from the target species, therefore making such tests more representative for toxicity assessment.
In addition to dose dependence, soil particle size, mineralogy, Pb speciation, and food type are factors that influence Pb bioavailability (Steele et al., 1990). Processes regulating interactions between different metal species and their bioavailability values include solubility, adsorption, complexation, redox reactions, and biological uptake (Samiullah, 1990). Thus it is critical to know Pb speciation to predict its bioavailability. Traina and Laperche (1999) reported that the toxicity of a metal is directly proportional to the activity of the free ion, and the most toxic solid phase of a contaminant will be that which supports the largest equilibrium activity of the metal. Thus, Pb minerals with the lowest solubility in the gastrointestinal tract will be the least poisonous to waterfowl. Bioavailability of Pb minerals has been assessed by PBET in several studies to determine which minerals contribute to higher metal bioavailability and pose the greatest toxicity potential (Davis et al., 1993; Ruby et al., 1994, 1999; Yang et al., 2001). Ruby et al. (1999) reported that Pb-mineral bioaccessibility increases in the order: galena < pyromorphite < Fe–Pb oxides < lead jarosite < Mn–Pb oxides < Pb oxides < cerussite.
The goal of this study was to develop a W-PBET method and correlate it with the bioavailability results obtained from a bird-feeding study (Heinz et al., 2004). The bioavailability test will be used to assess P-remediation potential for Pb-contaminated soils located in the Lower Coeur d'Alene River Basin. The soils are derived from fluvially deposited tailings, waste rock, concentrates, and smelter emissions that were transported by the South Fork of the Coeur d'Alene River. Such soils are generally defined as Slickens. The Pb in the ore body exists as galena; in the soils, however, the Pb is probably present as oxides, carbonates, and phosphates, and sorbed on Fe and Mn oxides. The contamination and characteristics of the soils are similar to numerous other Pb-mining-contaminated soils around the world. At such sites, there is a clear need to have a simple and inexpensive method to assess risks to waterfowl.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
2, but the pH of gastric content is usually higher because the secretions are diluted by ingesta (Sturkie, 1976). Intestinal pH ranges from about 5.2 to 7.2, and increases posteriorly due to pancreatic secretions and buffers secreted by the intestinal epithelium (Klasing, 1998). The average pH values of 2.6 (stomach phase) and 6.2 (intestine phase) were used in the W-PBET model. The effect of pH on metal extractability in the simulated gizzard for pH values of 2.0, 2.6, and 3.2 were measured.
Soil Mass and Fluid Volume
For the PBET model, the soil/fluid ratio of 1/160 (kg L–1) was selected so that diffusion-limited dissolution kinetics would not control the test results (Ruby et al., 1996). Rodriguez et al. (1999) applied a 1/150 soil/fluid ratio in the in vitro GI method to estimate bioavailable As in contaminated soils. Hamel et al. (1998) reported that the bioaccessibility of metals in the soils extracted by the in vitro synthetic gastric juice will only be slightly affected by changes in gastric fluid/solid ratios for the range 1/100 to 1/5000. In our study, the solid/solution ratio was chosen based on the waterfowl's daily ingestion of soil, which was derived from in vivo studies (Heinz et al., 2004), and 50 mL of gastric solution, which was the estimate for gizzard volume used by Levengood and Skowron (2001). A non-breeding mallard duck eats 70 to 100 g of food on a dry-weight basis per day, and the sediments constituted 12% of the diet (Heinz et al., 2004). Therefore, the ducks ate approximately 8.4 g of soil per day, and the estimated soil/solution ratio was 8.4 g to 50 mL = 0.168 kg L–1 (approximately 1/6). However, because the soil ingestion rate and retention time is variable in a living organism, the true soil/solution ratio is highly variable. To assess the effect of soil/solution ratio on W-PBET metal bioaccessibility, several ratios were tested.
Stomach Mixing
Regular rhythmic contractions in the gizzard create mixing and grinding (enhanced by the presence of small rocks). Peristaltic and segmenting movements comprise the mixing behavior in the bird's intestine (Sturkie, 1976). Mixing in the Ruby et al. (1993) PBET model was achieved by passing Ar at 1.0 L min–1 through the reaction mixture. Rodriguez et al. (1999) used individual paddle stirrers at a speed of approximately 100 rpm. We used a water bath shaker oscillating at 250 rpm to simulate mixing.
Soil Particle Size
Development of PBET models for humans considered soil particle sizes <250 µm because particles of this size would adhere to a child's hands, creating a route of ingestion (Duggan et al., 1985; Rodriguez et al., 1999; Ruby et al., 1996). Several ongoing studies use soils with particle size <500 µm to measure bioavailability to ecological receptors such as the shrew and American robin (Ruby, 2003). In our study, soil particles <1 mm were used because this is the size fraction used in the waterfowl feeding study (Heinz et al., 2004). Smaller particles have greater ratios of surface area to volume, hence, are more rapidly solubilized, which may result in greater Pb bioavailability (Ruby et al., 1992; Sparks, 1989). Waterfowl may contain grit (average is 45 g) in the gizzard (Klasing, 1998), which facilitates grinding. Grinding is not simulated in the centrifuge tubes when they are mixing in the water bath in the W-PBET test. Therefore, the effect of particle size on bioaccessibility was tested in this study.
Stomach Emptying Rate and Small Intestinal Transit Time
Rodriguez et al. (1999) concluded that the length of time to perform the stomach phase and intestinal phase for humans was not clearly described in the literature. In their study, they found that As concentrations in samples taken every 60 min remained constant for 3 h.
The length of time that food materials spend in the gizzard depends on their size. Small particles and liquid components pass through in minutes, whereas hard grains may remain in the gizzard for several hours. Klasing (1998) gives typical mean retention times in birds (the average time required for digesta to move through the GI tract); the mean retention time for herbivore birds is 50 to 300 min. In petrels, chickens, turkeys, and geese, about 50% of the mean residence time is spent in the stomach. Using these estimates, stomach incubation time is estimated to be 25 to 150 min. In our study, extraction times of 25, 60, and 150 min were tested.
Temperature
Because dissolution reactions are dependent on temperature, we expect a significant temperature influence on the rate and equilibrium status of the reactions occurring in the extraction test. In the PBET, the temperature is set to mimic a human (37°C). The waterfowl body temperature is 42°C, which was used in the W-PBET in our experiment (Levengood and Skowron, 2001).
Gastrointestinal Fluids
Levengood and Skowron (2001) examined concentrations of heavy metals in the gizzard contents of 18 mallards. Gizzard contents were transferred into 50 mL of simulated gastric juice containing 1 M NaCl, 10 g L–1 of pepsin, and HCl to adjust the pH to 2.0. These are the gizzard fluid parameters used in the W-PBET model, except pH was adjusted to 2.6. We applied the same intestinal solution containing bile salts and pancreatin as in the in vitro model by Rodriguez et al. (1999). Across species, the small intestine is considerably less variable than other organs because the diverse physical constitution of different foods is reduced to a relatively uniform fluid suspension, or chyme, by the action of the proventriculus and gizzard (Klasing, 1998). However, it is possible that different bile concentrations and bile salts of either porcine or bovine origin may induce different bioaccessibility values for the different models (Oomen et al., 2002).
Soils
The contamination sources are ore deposits of Pb–Zn–Ag veins occurring in Precambrian rocks of the Belt Supergroup. Minerals associated with the ore bodies are quartzite, argillite, pyrite, quartz, and carbonates such as siderite. The principal economic minerals are galena (PbS), sphalerite (ZnS), and argentiferous tetrahedrite [(Cu,Fe)12Sb4S13]. As a result of milling and weathering, Pb phases in the soils are probably secondary mineral species (oxides, phosphates, and carbonates), and associated with Fe and Mn oxides. The soils are high in Fe and Mn oxides, containing on average 7% Fe and 0.4% Mn by weight, which are nearly all secondary oxide mineral phases (Hickey and Strawn, 2004). As a result of the presence of secondary oxides, Pb most likely exists as Pb–Fe oxides, Fe–Mn oxides, or both within the soils.
Soil samples used in measuring method reproducibility and accuracy and sensitivity analysis were collected from a soil remediation test site located on the northwest shores of Bull Run Lake in the Lower Coeur d'Alene Basin. Total Pb concentration in this soil is 5300 mg kg–1. The same soil samples used in the mallard-feeding study (Heinz et al., 2004) were used in this study. The study site includes a control plot (no treatments) and a P-amended plot. Unamended soils collected from the control plot were used in method validation. Samples were collected using a 20-cm-long 5-cm-diameter stainless steel sampler with a plastic sleeve insert. Core samples were submersed in liquid N2, sealed, placed on ice, and transported back to the laboratory where they were kept at –5°C. Soils were air dried and gently crushed to pass a sieve of 1-mm pore size. Total metal concentration was measured on the soils using a HF/aqua regia digest in a Teflon microwave digestion cell as outlined in EPA Method 3052 (USEPA, 1991).
Soils fed to mallards in the mallard-feeding study (Heinz et al., 2004) were used in the W-PBET model to investigate relations between in vitro and in vivo Pb. Lead-contaminated soil samples from the Coeur d'Alene River Basin (Harrison Slough, Black Rock Slough, and Bull Run Lake soils) in Idaho were P amended in either laboratory incubations or field trials (Heinz et al., 2004, Table 2). A soil sample from Round Lake (in the St. Joe River watershed in Idaho) that had low Pb concentrations was used as an uncontaminated control soil.
|
Waterfowl Physiologically Based Extraction Test Procedure
We used a two-step sequential extraction consisting of the gastric and intestinal phases as separate measurements of gastrointestinal availability. This approach is similar to the PBET model developed by Ruby et al. (1996), yet modified for a waterfowl's physiology as described above. Blanks (no soil) and the standard soil were run through the experiment.
Gizzard Phase
The gizzard solution consisted of 1 M NaCl and 10 g L–1 pepsin (from porcine stomach mucosa, Sigma Chemical Co.,) acidified to pH 2.6 with HCl (Kimball and Munir, 1971). Thirty milliliters of the gizzard solution were combined with 3.6 g of contaminated soil in a 50-mL polycarbonate centrifuge tube. The tube was degassed with high-purity N2 (g), sealed, and placed in a water bath at 42°C. Samples were mixed in the water bath at 250 rpm. All solutions were equilibrated for pH and temperature before adding to the soil. Following 1 h of incubation, the samples were removed and centrifuged, pH was measured, and the sample was filtered. Measurements of pH taken before and after centrifuging were not significantly different; therefore, pH was measured only after centrifuging. Samples were centrifuged for 24 min at 2404 relative centrifugal force, and the supernatant was filtered through a 25-mm syringe filter with a 0.2-µm membrane (Gelman Laboratory, Ann Arbor, MI) following pH measurement. Because the proteins and salts may cause a high background effect and clogging during analysis, the samples were diluted 1:10. The filtrate was analyzed for Pb using ICP-AES. A multi-element standard (CPI International, San Antonio, Texas) was used to assure accuracy in measuring Pb concentrations on ICP-AES. Detection limits for both the gizzard- and intestine-phase Pb on the ICP-AES were 0.01 mg L–1.
Intestinal Phase
Following the gizzard phase, a separate set of uncentrifuged samples were adjusted to pH 6.2 by adding the appropriate amount of NaHCO3 saturated solution (1 mL). Bile salts and pancreatin (from porcine pancreas, Sigma–Aldrich Co., St. Louis, MO) were added in the amount of 0.35% (0.105 g per 30 mL) and 0.035% (0.0105 g per 30 mL), respectively (Rodriguez et al., 1999). Samples were mixed in the water bath at 250 rpm. Following incubation for 2 h, the samples were removed and centrifuged, pH was measured, and the sample was filtered. All other aspects of the sample treatment and analysis were the same as the gizzard phase.
Effect of Test Parameters on Metal Extractability
Sensitivity analysis was conducted on the W-PBET model (gizzard phase) to determine the effects of pH, grinding, soil/solution ratio, and extraction time on Pb extractability in the simulated gizzard. Discussion of the relevant range for these parameters is provided above. Bull Run unamended soil samples were run at pHs of 2.0, 2.6, and 3.2 through the W-PBET gizzard phase. Another W-PBET experiment was conducted on soils to test extraction times of 25, 60, and 150 min. All other parameters were as described in Table 1. Grinding effect was tested on soils with particle sizes of <1 and <0.25 mm. The <1-mm size fraction was ground to pass through the 0.25-mm sieve. Soil/solution ratio effect on W-PBET metal bioaccessibility was investigated at 1/6, 1/8.3, 1/100 and 1/200 g/mL. All samples were run in triplicate.
Data Analysis
The W-PBET metal bioaccessibility was calculated as the ratio of metal concentration in the extracted phase (in milligrams per kilogram) to total Pb concentration in the soil (in milligrams per kilogram). Absolute Pb bioavailability was calculated from results of the in vivo experiment tissue Pb concentrations (in milligrams per kilogram, wet weight) (Heinz et al., 2004) divided by total Pb concentration in the diet (in milligrams per kilogram) (USEPA, 1999).
Statistical analyses were performed using SAS version 8.2 (SAS Institute, 2001). Least significant difference t-tests were used to separate means. A Spearman correlation test was used to determine if W-PBET Pb was related to mallard feeding test results. Linear regression was performed for metal gizzard extractabilities and W-PBET parameters.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Sensitivity Analyses
Simulated gizzard pH is one of the governing factors of Pb extractability. Results for Pb extractability in the simulated gizzard at different pHs indicated that there were linear relations between pH and Pb concentration in the gizzard phase in the pH range of 2 to 3.2 (regression coefficient R2 = 0.97) (Fig. 1). Lead concentration in the gizzard extraction doubles as pH decreases from 3.0 to 2.0. Because the pH of a bird stomach is variable, depending on food presence and bird species, we chose the mid-range pH value of 2.6 to use in the W-PBET experiments. However, the linear relationship between pH and extractable Pb shows that, while pH will impact the absolute bioaccessibility value, the correlation between W-PBET Pb and in vivo Pb should remain the same.
|
|
|
|
|
|
|
The Harrison Slough soil had greater bioaccessible Pb values than the other soils while, in the mallard-feeding test, the differences between the Harrison Slough soil and the other soils were much less. All of the soils had a similar particle size distribution (Bull Run Lake soils are very fine sandy loam and Black Rock Slough and Harrison Slough soils are silt loams), eliminating particle size as the reason for the differences in the W-PBET and in vivo results. Thus we conclude that the different results for the two methods are due to Pb speciation, soil mineralogy, or the presence of another component (e.g., more available P, soil organic matter, etc.) or some combination of these factors. Without doing speciation studies, we cannot further speculate on the reason for the observed differences in toxicity assessment between the two methods for the Harrison Slough sample.
Furman (2004) showed that reduced soils have different metal bioaccessibilities than oxidized soils, which could only be due to soil mineralogy changes on oxidation (Pb is not redox sensitive under environmental conditions). Soil mineralogy can impact the amount of Pb in solution because different minerals will maintain different solution concentrations of Pb in the simulated digest solution (Ruby et al., 1999). This difference highlights the importance of understanding how soil variability can influence bioavailability, and the need for tests that can measure such variation.
Bioaccessibility of Lead in Phosphorus-Amended Soils
The W-PBET results, as well as mallard-feeding results, indicate that P amendments significantly reduced Pb bioaccessibility and bioavailability (Table 2). The results of Heinz et al. (2004), however, suggested that reduced bioavailable Pb in the soils amended with 1% P would still present hazards to waterfowl. Soils from different locations in the Lower Coeur d'Alene Basin had different W-PBET Pb values, suggesting that the speciation is different at the different sites (Fig. 5a and 6). Maenpaa et al. (2002) observed different reduction values in metal bioavailability to earthworms in P-amended soils from different locations and concluded that soil characteristics other than P amendments affect metal bioavailability to earthworms. According to W-PBET results, Pb showed a significant reduction in bioaccessibility in all P-amended soils (Fig. 6). This indicates that P amendments immobilized Pb species.
Many studies have investigated the decrease in Pb bioavailability in P-amended soils and have proposed that formation of Pb phosphates, such as pyromorphites, are responsible for the reduced bioavailability (Hettiarachchi et al., 2000; Laperche et al., 1997; Melamed et al., 2003; Ruby et al., 1994). In solutions with several metals present, the phase with the lowest solubility will usually precipitate from solution before the more soluble metals (Cao et al., 2003). Compared with other metals (such as Cd, Zn, and Mn), Pb forms the least soluble minerals (Lindsay, 1979). Pyromorphite formation is a quick process when ions are in solution (Scheckel et al., 2003). However, the availability of P and Pb in the soil solution to form new phases also needs to be considered. For example, Pb and P may be adsorbed on Fe oxide surfaces, making them unavailable for precipitation reactions.
Reactions between Pb and P can also take place in the acidic conditions of the simulated gastrointestinal fluid and result in formation of pyromorphite in vivo (Zhang and Ryan, 1998). Lead phosphates generally have low solubility; however, Oomen et al. (2003) used voltammetry to measure Pb species in solution and observed that Pb phosphate complexes are soluble in the simulated chyme of the human digestive system and could therefore be a source of Pb that is available for transport across the intestinal epithelium. This suggests that poorly soluble Pb phosphate minerals may not be completely unreactive in the digestive system.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Although W-PBET was designed to simulate the waterfowl digestive tract, it must be emphasized that extractable Pb should be viewed in a relative context. The use of the predicted Pb bioaccessibility for absolute bioavailability predictions is weakened by the assumptions within the W-PBET model and the difficulty in precisely simulating bio-uptake in the GI system. Further tests on how soil physicochemical properties impact bioaccessibility will be useful to further refine the model.
Because the geochemistry of Pb and other metals in the soils is dynamic, it is critical to have an assessment tool that will allow scientists, managers, and engineers to evaluate how environmental variables and remediation and management strategies might impact Pb bioavailability. The W-PBET model is a relatively cost effective, reproducible, and easy method that can be used for this purpose. Validation of the model on a broader range of soils is required, however, before it can be universally adopted as a tool for Pb bioaccessibility assessment in waterfowl.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  O. Furman, D. G. Strawn, and S. McGeehan Sample Drying Effects on Lead Bioaccessibility in Reduced Soil J. Environ. Qual., May 7, 2007; 36(3): 899 - 903. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal |
