SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, M.
Right arrow Articles by Harris, W. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Chen, M.
Right arrow Articles by Harris, W. G.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Chen, M.
Right arrow Articles by Harris, W. G.
Related Collections
Right arrow Other Soil Types
Right arrow Heavy Metals
Soil Science Society of America Journal 66:632-640 (2002)
© 2002 Soil Science Society of America

DIVISION S-8—NUTRIENT MANAGEMENT & SOIL & PLANT ANALYSIS

Arsenic Concentrations in Florida Surface Soils

Influence of Soil Type and Properties

M. Chen*,a, L. Q. Mab and W. G. Harrisb

a Institute of Geography and Natural Resources, Chinese Academy of Sciences, Beijing 100101, China
b Soil and Water Science Dep., University of Florida, Gainesville, FL 32611-0290

* Corresponding author (mchen{at}mail.ifas.ufl.edu) and current address: Everglades Education and Research Center, University of Florida, 3200 E. Palm Beach Road, Belle Glade, FL 33430


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 
Background As concentrations in soils are important for defining whether a soil is polluted. Arsenic concentrations in 441 taxonomically and geographically representative surface soils were determined using EPA Method 3052 (HCl-HNO3-HF digestion). Cumulative distribution plots indicate that As concentrations follow a log-normal distribution and depend on soil type. Sample geometric mean(GM) (the exponential mean of the log-transformed distribution) As concentrations (mg kg-1) generally follow the soil taxonomic order of Histosols (2.35) > Inceptisols (0.98), Mollisols (0.72) >= Ultisols (0.51) >= Alfisols (0.39), and Entisols (0.36) > Spodosols (0.18). The highest As concentrations were found in soils that occur exclusively or prevalently in wetlands, such as Hemists (3.16–9.44), Saprists (0.15–11.7), Aquents (0.10–50.6), Aquolls (0.03–3.34), and Aquepts (0.03–38.2). Both linear and multiple regressions indicate soil properties (clay, pH, cation-exchange capacity [CEC], organic C, and total Al), especially total Fe and P, are important factors affecting natural background concentrations of As in Florida soils. Arsenic release from bedrock (limestone) and As bioaccumulation by aquatic organisms are possible explanations for relatively high As in those wetland soils. The use of a single regulatory value criterion for As contamination in soil cannot provide an adequate assessment given the natural variation in soil As. Baseline soil-As concentration, which was defined as 95% of the expected range of background As concentrations in different soil categories, is necessary for properly assessing potential As contamination.

Abbreviations: CEC, cation-exchange capacity • ENP, Everglades National Park • FDEP, Florida Department of Environmental Protection • GM, geometric mean • GSD, geometric standard deviation • NJDEP, New Jersey Department of Environmental Protection • UBC, upper baseline concentration


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 
PUBLIC CONCERNS OVER AS POLLUTION in both soils and waters have substantially increased in recent years (Association for the Environmental Health of Soils, 1998; Research Triangle Institute, 1998). Long suspected to be responsible for bladder, kidney, liver, lung, and skin cancers, inorganic As has been listed by the USEPA as a Class A human carcinogen (Research Triangle Institute, 1998). The current maximum allowable levels of As for oral intake and drinking water are set at 0.3 µg kg-1 d-1 and 50 µg L-1, respectively (USEPA, 1998). However, the World Health Organization has recently begun to press for stricter consumption standards and has recommended lowering the As drinking water standard to 10 µg L-1 (Research Triangle Institute, 1998). The USEPA is currently under pressure from the National Research Council to lower its current drinking water standard of 50 µg As L-1 to <20 µg As L-1 to adequately protect public health (Christen, 1999).

In addition to concerns over As contamination in water sources, it is also known that dietary intake of As through the food chain via uptake from contaminated soils may adversely affect human health (Arnt et al., 1997). While the regulations governing As contamination in waters are well defined, the regulatory cleanup goals for remediation of contaminated soils are still under development and may vary greatly among countries, states, and land uses. For example, the regulatory limits set by the Ministry of Environment of Canada for As contamination in agricultural, industrial, and residential soils are 25, 50, and 25 mg kg-1, respectively (Arnt et al., 1997). In contrast, those for the United Kingdom are set at 10 mg kg-1 for domestic gardens and 40 mg kg-1 for parks, playing fields, and open spaces (O'Neill, 1990). Variations of this type are also found in the U.S. regulatory guidelines (Association for Environmental Health of Soils, 1998). The New Jersey Department of Environmental Protection (NJDEP) has set a state cleanup criterion of 20 mg kg-1 for both residential and nonresidential soils based on background soil As levels (Barringer et al., 1998; NJDEP, 1998). The current soil cleanup goals set by the Florida Department of Environmental Protection (FDEP) for As in residential and industrial soils are 0.80 and 3.7 mg kg-1, respectively (Tonner-Navarro et al., 1998), based on direct exposure. Variations in these guidelines indicate a need for standardizing how cleanup criteria are established. In cases where the risk-based criteria are lower than the method detection limit (NJDEP, 1998) or are below the site-specific background level, it is reasonable to use the latter two as the cleanup criteria (Association for Environmental Health of Soils, 1998). Therefore, it is important to obtain accurate As background concentrations for different soils to prevent unrealistically low mandatory guideline levels (Kabata-Pendias and Pendias, 1992; Davies, 1992). To obtain the accuracy required for regulation purpose, As background levels in soils should be based on a sufficiently large database. These data can then be used as a reference to compare with site-specific As concentrations (Breckenridge and Crockett, 1995) in determining whether or not a particular soil type is contaminated.

The main source of As in soils is the parent materials from which the soil is derived. Atmospheric deposition contributes significantly to the geochemical cycle of As (Smith et al., 1998). Also, bioconcentration of As by lowland plants and aquatic organisms, such as algae, seagrass (Halophila Thouars), and lower invertebrates (Otte et al., 1990; Cullen and Reimer, 1989; Strom et al., 1992) contributes to elevate As concentrations in lowland soils. Doyle and Otte (1997) found that salt marshes could act as effective sinks for As and other metals. In both soil and water systems, As species are subject to chemical and microbiological oxidation-reduction reactions according to the EH-pH status (O'Neill, 1990; Masscheleyn et al., 1991). At high EH values (upland), As (V) may predominantly exist as H2AsO-4 (pH = 2.5–7.0) and HAsO2-4 (pH > 7.0), constituting >90% of the total soluble As. At low EH values (lowland or wetland), As (III) species (H3AsO3, pH < 9.0) may be present along with As2S3 (pH < 7.0) and AsS-2 (pH > 7.0) if there is sufficient S present (Masscheleyn et al., 1991).

Florida soils form primarily from well-weathered sandy marine sediments and contain little weatherable primary minerals. The small amount of clay generally present commonly consists of secondary minerals such as kaolinate, hydroxy-interlayered vermiculite, gibbsite, and quartz, as cemented by lesser amounts of metal oxides (Harris et al., 1995). Seven soil orders have been identified in Florida; their names and approximate percentages are as follows: Spodosols (28%), Entisols (22%), Ultisols (19%), Alfisols (14%), Histosols (10%), Mollisols (4%), and Inceptisols (3%) (Chen et al., 1999). Efforts have been made to establish As background concentrations in noncontaminated soils in Florida. A level of 8 mg kg-1 in noncontaminated agricultural soils has been reported and frequently cited (Smith et al., 1998). Scarlatos and Scarlatos (1997) reported that As concentrations in 115 surface soil samples collected near Homestead, Florida ranged from 1.1 to 54.3 mg kg-1, exceeding the FDEP residential soil cleanup goal of 0.80 mg kg-1 (Tonner-Navarro et al., 1998). A GM (exponentiated mean of a log-transformed distribution) As level of 1.1 mg kg-1 in surface soil horizons has also been reported based on 40 Florida mineral soils (Ma et al., 1997). Recently, a comprehensive study to evaluate baseline concentrations of 15 potentially toxic trace elements (Ag, As, Ba, Be, Cd, Cr, Cu, Pb, Hg, Mn, Mo, Ni, Sb, Se, and Zn) was conducted and a relatively lower GM of As (0.42 mg kg-1) was noted (Chen et al., 1999). This investigation lays a general foundation for understanding baseline soil concentrations of a variety of trace metals including As at the state level. However, no detailed information is available regarding how the As concentration variation associates with the soil classification system and is influenced by soil properties.

The present study was conducted (i) to determine As concentration distribution characteristics associated with selected levels of soil taxonomy, and (ii) to examine soil factors related to As levels in soils. It is hoped that such information will help to evaluate the significance of anthropogenic and natural sources of As inputs in soils by establishing normal As concentration ranges in different soil types. In addition, correlation between As concentrations and soil properties is useful in assessing As contamination by defining soil background concentration in a specific area using a regression model.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 
Soil Sample Selection and Analysis
This study was based on a previously established database (Chen et al., 1999) which contains As concentrations of 448 taxonomically and geographically representative, air-dried archived Florida surface soils. Those soils were previously sampled as a part of the Florida Cooperative Soil Survey Program conducted jointly by the Soil and Water Science Department of the University of Florida and the NRCS–USDA during 1967 to 1989 (Sodek et al., 1990). Soil horizons were identified and sampled using the USDA soil survey conventions and procedures (Soil Survey Staff, 1993). Bulk samples were collected from each horizon, air-dried, and stored in cylindrical cardboard cartons at ambient temperature in a special building at the University of Florida Soil Characterization Laboratory. Only surface soil horizons were used in this study and aliquots of stored samples were obtained with special permission. The selected soil samples in the current study were from rural areas and were originally selected based on their representation of the soil series being mapped. Soils were from 51 of 67 counties and their map units covered ~80% of the total land area of Florida (Chen et al., 1999). Most of the selected sites (60%) were under native vegetation at the time of sampling, although some had been influenced by agriculture. Approximately 40% of the samples were described as having surface horizons that had been disturbed either by plowing or clearing.

Because of its complexity, the database was divided into several more manageable levels of generalization based on the soil classification scheme. For example, soil suborders with <5 samples were excluded from the database. As a result, only 441 out of the 448 soil samples in the database of Chen et al (1999) were used. These 441 soil samples included seven soil orders and twelve soil suborders: Histosols (Hemists and Saprists), Inceptisols (Aquepts), Mollisols (Aquolls), Ultisols (Aquults and Udults), Entisols (Aquents and Psamments), Alfisols (Aqualfs and Udalfs), and Spodosols (Aquods and Orthods). All Histisols and some Aquic (wet) suborders (Aquents, Aquepts, and Aquolls) were grouped as wetland soils, whereas all Spodosols and Psamments, Udalfs, and Udults suborders were grouped as upland soils.

Soil samples were handled with caution and ground to pass a 2-mm plastic sieve using an agate mortar and pestle to ensure that trace metal contamination did not originate from the grinding process. One gram of each soil sample was weighed and digested in a CEM MDS-2000 microwave oven (CEM, Matthews, NC) using modified EPA Method 3052 (USEPA, 1995). Quality assurance samples (a blank, a duplicate, a spike, and a standard reference material) were included for every 20 samples in the digestion following an FDEP approved research quality assurance plan (Chen, 1997). Arsenic concentrations in the digestates were analyzed on a Perkin-Elmer SIMAA 6000 graphite furnace atomic absorption spectrophotometer (Perkin-Elmer, Norwalk, CT) with Zeeman background correction using EPA SW 846 method 7060A (USEPA, 1995). The method detection limit for As is 0.03 mg kg-1 soil. Concentrations of Al, Fe, and P were determined with an inductively coupled plasma spectrophotometer (Thermo Jarrell Ash ICAP 61-E. Analytical Instrument Recycle, Inc., Golden, CO) using EPA SW 846 method 6010 (USEPA, 1995). Particle-size distribution, organic C, and CEC were previously determined by a modified pipette method, by the Walkley-Black acid dichromate digestion method (soil survey procedure No. 6A1c), and by the sum of extractable bases (Ca, Mg, Na, and K) and extractable acidity (soil survey procedure No. 6H4a), respectively (Sodek et al., 1990).

Statistical Analysis
All statistical analyses were performed using a PC version of the SAS program (SAS System 6.11, SAS Institute, Gary, NC). Analysis of variance and a simple paired t-test at a confidence level of {alpha} < 0.05 was used to assess significant differences between soil orders or suborders. Simple correlation analysis was used to determine method correlation coefficients for each element at probability levels of {alpha} < 0.05, 0.01, and 0.001, respectively (SAS Institute, 1987). A stepwise multiple regression analysis was used to model the As data using clay, pH, organic C, CEC, and total concentrations of Al, Fe, and P: Y = ß0 + ß1 X1 + ß2 X2 + ... + ßp Xp + {epsilon}, where ß0 denotes the intercept, ßi (i = 1,2, ..., p) denotes the slope for the variable Xi (Millard and Neerchal, 2001). Partial correlation coefficients were also calculated to show the contribution of individual variables to the total explained variance.

A cumulative frequency distribution is a fundamental graphic format used to display the shape of a distribution by plotting the number of observations from a sample data set falling in or below an interval (Myers, 1997). In the current study, a cumulative frequency histogram was used to test the As distribution normality in soils. The x-axis plots a histogram of the data categories; the y-axis plots the cumulative frequency of the data in percentage. Cumulative frequencies will plot a straight line if the distributions are normal (Myers, 1997). Rather than straight lines, the plots showed markedly convex shapes, highly characteristic of positively skewed data (Fig. 1) . The data sets were then normalized by log 10 transformation prior to statistical comparison (Breckenridge and Crockett, 1995) unless specified otherwise. Geometric mean, which was calculated as the exponentiated mean of the log-transformed data, was used to estimate the true median for the lognormal distribution. Geometric standard deviation (GSD), which was calculated as the exponentiated standard deviation of the log-transformed data (Millard and Neerchal, 2001), was also used to characterize the lognormal distribution. Baseline As concentration in soil was defined as 95% of the expected range of background concentration and calculated as GM/GSD2 and GM x GSD2 (Dudka, 1993; Gough, 1993). The upper baseline concentration (UBC), which was defined as the 97.5% of the lognormal distribution and calculated as GM x GSD2 (Chen et al., 1999), was used to evaluate soil samples for possible As contamination.



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 1. Cumulative probability curves for As concentrations in Florida surface soils. Numbers in parentheses are sample sizes of individual soil orders.

 

   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSION
 REFERENCES
 
Distribution Characteristic of Arsenic Concentrations in Different Soils
Cumulative probability distributions of As concentrations in seven major Florida soil types are shown in Fig. 1. The shape of the curve represents the distribution characteristic of As for each soil order, with steeper slopes indicating a narrower spread of concentrations and horizontal offsets indicating greater differences in As concentrations (Peters et al., 1999). Arsenic concentrations varied significantly among different soil types in Florida. Generally, As concentrations decreased in the order of Histosols > Inceptisols, Mollisols > Ultisols > Entisols, Alfisols > Spodosols (Fig. 1, Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Arsenic concentrations in different types of Florida surface soils based on soil orders and soil suborders.

 
More than 95% of the Histosols (all except one) had As concentrations above the current USEPA soil screening level of 0.40 mg kg-1 (USEPA, 1996) while only 23% of the Spodosols were above this limit (Fig. 1). In addition, only 6% of the Spodosols, 25% of the Entisols, Alfisols, and Ultisols had As concentrations above the current FDEP residential soil cleanup goal of 0.80 mg kg-1 compared with ~50% of the Mollisols and Inceptisols and 90% of the Histosols. Fewer than 5% of the Spodosols, Alfisols, Mollisols, and Ultisols were above the FDEP industrial soil cleanup goal of 3.70 mg kg-1, compared with 10% of the Entisols, 25% of the Inceptisols, and 30% of the Histosols. None of the Mollisols, one Alfisol, Inceptisol, and Spodosol, and two Ultisols had As concentrations above the UBC of 7.02 mg kg-1 for Florida surface soils reported by Chen et al. (1999), compared with 7% of the Entisols and 15% of the Histosols (Fig. 1). In general, organic soils (Histosols) had greater As concentrations than mineral soils.

It is interesting to note that the two suborders of Entisols, Psamments (the most prevalent) and Aquents, differed markedly in their As content with the latter generally having greater As concentrations (Table 1). This is because the Aquents (marl soils) are dominated by silt- and clay-sized calcium carbonate (4.2–94%, with an average of 56%), which are common in south Florida. Anthropogenic inputs of As to marl soils were suspected since most marl soils had been influenced by agriculture. Arsenic concentration of up to 60 mg kg-1 in organic-rich sediments from south Florida has been documented by the U.S. Geological Survey to address the general issue of south Florida wetland degradation (Gough et al., 1996). Therefore, we analyzed As concentrations in an additional eleven newly collected marl soil samples (all are Aquents) in 1998 that were not in the 448 soil database, with five from the Everglades National Park (ENP) and six from other undisturbed areas of the region. We found that As concentrations in those soils were also relatively high. Arsenic concentrations in the ENP samples ranged from 2.90 to 24.9 mg kg-1, with a GM of 5.37 mg kg-1, and those of the other undisturbed marl soils ranged from 3.10 to 13.3 mg kg-1, with a GM of 4.70 mg kg-1. This result seems to indicate that soils in that specific region could be naturally high in As concentrations (Chen et al., 2000). Other studies have reported similar As enrichments in natural soils or sediments in that area (Scarlatos and Scarlatos, 1997; Strom et al., 1992; Gough et al., 1996). It has been reported that the magnitude of the sorption maximum for calcite is four times that for kaolinite and montmorillonite (Goldberg and Glaubig, 1988). In the cases of pH above 7.5 (Aquents), carbonates may play a major role in As sorption. The relatively great concentrations of As and other elements in the marl soils may also attribute to elemental accumulation during formation of calcite (Chen et al., 2000).

The highest mean As concentrations were associated with Histosols, especially the suborder of Hemists (Table 1). The Hemists had a GM As concentration of 5.60 mg kg-1, which is greater than the FDEP industrial soil cleanup goal of 3.70 mg kg-1. Even when bulk density is taken into account and the As concentrations are compared on a volume basis, the Histosols still have a higher GM As concentration. This is expected since the Histosols had the greatest values of organic C, total Fe, Al, P, and CEC (Table 2). Statistical analysis showed no difference in GM As concentrations between the disturbed (n = 10) and undisturbed (n = 29) Histosols from the same region (2.37 vs. 2.34 mg kg-1) (data not shown). Thus, the relatively high As concentrations in the Histosols, especially in the Hemists, apparently relate to natural biogeochemistry rather than from arthropogenic sources.


View this table:
[in this window]
[in a new window]
  		Table 2. Properties of different Florida soils and their correlation with As concentrations based on soil orders.

 
Inceptisols and Mollisols have the second and third highest As contents, with GM As concentrations of 0.98 mg kg-1 and 0.72 mg kg-1, respectively (Table 1). These two soil orders have the highest clay contents (53 g kg-1 and 3.8 g kg-1, respectively). Additionally, the Inceptisols had the highest CEC (17.4 cmol kg-1) excluding the Histosols (129 cmol kg-1) and the Mollisols had the highest organic C, pH, and total Fe, Al, and P concentrations among the mineral soils (Table 2). These results suggest the importance of clay and total Fe, Al, and P in controlling As levels in soil. Ori et al. (1993) reported that total As in Louisiana soils appeared to vary more with the parent material and clay content than with the soil reaction or organic matter content. The Inceptisols (Aquepts) and Mollisols (Aquolls) in Florida were mainly wetland soils, which may undergo different biogeochemical processes responsible for As sorption compared with the upland soils (McGeehan et al., 1998).

Ultisols had the fourth highest overall GM As concentration (0.51 mg kg-1) and the highest among the three dominant mineral soil orders in Florida (Spodosols, Entisols, and Ultisols). This finding agrees well with Ma et al. (1997), who found that As concentrations in 40 Florida mineral soils decreased in the order of Ultisols > Entisols >= Spodosols. The relatively high concentrations of total Al, Fe, and P in the Ultisols and Entisols compared with that found in the Spodosols (Table 2) could be the primary reason.

Baseline Arsenic Concentrations in Soil Taxonomy
Baseline soil concentration has been proposed as a reference to represent elemental concentrations specific for a given region and time period to determine clean soils and was recognized as the only means to establish reliable natural background concentration in soils worldwide (Chen et al., 1999). However, because of the significant differences in both cumulative frequency distribution and GM As concentrations among soil orders and suborders, it is inappropriate to use a single number to represent As concentrations in Florida surface soils. Baseline soil-As concentrations are therefore necessary to be defined separately, based on soil taxonomy and magnitude of the variation in As concentrations, for properly assessing potential As contamination in different soils. For example, an As concentration above the UBC of 7.02 mg kg-1 could occur naturally for Hemists and Saprists (Histosols), Aquents (Entisols), and Aquepts (Inceptisols) but may indicate potential As contamination for other studied soil orders and suborders (Table 1).

A careful examination of soil samples with As above 7.02 mg kg-1 revealed 12 of 17 those soil samples had very high contents of either clay (123–609 g kg-1, with an arithmetic mean of 311 g kg-1) or organic C (209–542 g kg-1, with an arithmetic mean of 386 g kg-1) (Tables 1 and 3). The 12 soil samples were all from wetland soil types and included four Sapists, two Hemists, five Aquents, and one Aquept. A recent study on the sawgrass (Cladium P. Br.) prairie wetlands in South Florida (U.S. Geological Survey, 1996) indicated that nonessential trace elements (such as Cr, Co, Pb, and Hg) are concentrated in organic-rich sediments. Bioaccumulation of As by aquatic organisms may play an important role in those soils with high As concentrations. Snail (Nassarius obsoletus) shells were observed in the marl soil sample (Aquents suborder) that had a concentration of 50.6 mg As kg-1. The soil sample containing the next highest As level (38.2 mg kg-1) was not a marl soil. It was a fairly sandy soil from a coastal hammock in Levy County, which contained mollusk shells. In fact, the common characteristic of these two relatively different wetland soils with the highest As concentrations was the abundance in shell fragments and the presence of underlying limestone. Other biological materials (debris) that could not be identified were also present. It has been reported that mollusks and crustaceans are capable of assimilating As from the water and concentrating it in their shells (Otte et al., 1990; Cullen and Reimer, 1989). For example, As concentrations in sea snails can be as high as 3.6 to 63 mg kg-1 on a wet-weight basis (Committee on Medical and Biological Effects of Environmental Pollutants, 1977). Elevated As concentrations in both mollusks and sediments from the south Atlantic and along the Gulf Coast of Florida have also been reported (Presley et al., 1990; Valette-Silver et al., 1999). Phosphorus deposits and soil pesticide residues were the hypothesized main sources of the elevated As, and the enrichment mechanism appeared to result from a mixture of processes including atmospheric deposition, river and aquifer inputs, and ocean upwelling (Valette-Silver et al., 1999). This is quite possible since phosphate rocks have a relatively high As concentration (6.6–121 mg kg-1) compared with the average (1.8–6.6 mg kg-1) nonphosphorite rocks (Van Kauwenbergh, 1997).


View this table:
[in this window]
[in a new window]
  		Table 3. Properties of soil samples with As concentrations greater than the upper baseline concentration of 7.02 mg kg-1 (Chen et al., 1999).

 
Arsenic Concentrations and Soil Properties
It has been reported that soil composition, such as contents of clay, organic C, and total Fe and Al, and soil properties, such as pH and CEC, are the major controlling factors for trace metal concentrations in soils (Chen et al., 1999). Geochemical associations of As with Fe and Al have been investigated by various researchers (Jacobs et al., 1970; Barringer et al., 1998) since As is strongly sorbed by Fe and Al oxides. Cullen and Reimer (1989) concluded that in both coastal marine and freshwater sediments, cycling of inorganic As is closely coupled with Fe biogeochemistry, which is effective in controlling As retention during flooding and drainage cycles. In the current investigation, contents of clay, organic C, total Fe, Al, and P, pH, and CEC in different types of Florida soils varied greatly (Table 2). Significant correlation existed between As concentration and total Al (r = 0.98), total Fe (r = 0.94), total P (r = 0.95) and clay content (r = 0.99), based on soil orders (Table 2). The capacity of a soil to retain As is strongly related to the abundance of Fe and Al oxides, particularly noncrystalline forms (Jacobs et al., 1970). The retention of As on clay is dependent on the quantity and type of clay-sized minerals. For example, kaolinite sorbs more As from solution than vermiculite, which sorbs more As than montmorillonite (Goldberg and Glaubig, 1988; Onken and Adriano, 1997). Organic matter generally has a low affinity for As. Humic substances in organic sediments can serve as strong reducing agents and can influence the processes that control mobilization of As (Gough et al., 1996). A significant correlation between As concentrations and soil organic C in the current study may indicate an important role played by organic C in As redox chemistry. Linear prediction equations could be used for total As concentrations vs. increasing clay, organic C, CEC, and concentrations of total Fe, Al, and P in all soils studied (Fig. 2) . The coefficient of determination between As concentration and soil property was 0.25 for clay, 0.19 for organic C, 0.24 for CEC, 0.46 for total Fe, 0.36 for total Al, and 0.39 for total P (n = 441, {alpha} < 0.001), respectively. Arsenic concentrations were also significantly correlated with CEC (r = 0.97) based on soil order (Table 2). This is possibly because CEC was affected by contents of clay, organic C, and total Fe, Al, and probably P in soils. Prediction equations are not significant ({alpha} < 0.05) and cannot be used for As concentration vs. pH. However, when contents of clay, organic C, CEC, pH, concentrations of total Al, Fe, and P were used in a stepwise regression to predict soil As concentrations, final models could be developed:

and



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 2. Arsenic concentrations vs. soil properties (contents of clay, organic C, CEC, total Fe, Al, and P) in the soils studied.

 
The calculated equations showed the considerable influence of soil pH and concentrations of total Fe and P in the prediction of soil As concentrations. In both regressions the partial correlation coefficients for total Fe and P were significant at {alpha} < 0.05. The high correlation between As concentration vs. total P is possible because phosphate deposits in Florida were also responsible for the high background P concentrations in soils (Chen and Ma, 2001). These findings also supported the hypothesis that total Fe was the most important soil factor affecting As concentrations in Florida soils. According to the stepwise multiple regression, contents of clay and organic C were not recognized as important soil properties that influences As concentrations. This is possible because of clay minerals present in Florida soils generally are coated by various metal oxides, such as Fe and Al (Harris et al., 1995). These coefficients also correspond with the results of Doyle and Otte (1997), who stressed oxidation of Fe, rather than the binding to organic matter, that drives the accumulation of As in soils.

Arsenic Enrichment and Retention in Wetland Soils
It has been reported that As background concentrations in noncontaminated soils rarely exceed 15 mg kg-1, except for the soils derived from weathered pyrites or quartzite (Smith et al., 1998). The soil suborders with the highest As concentrations in Florida tended to be wet. These included Saprists and Hemists, which are exclusively wetland soils, and Aquents, Aquepts, and Aquolls, which are typically found in wetlands. In addition, these soils are also commonly associated with limestone. For example, Saprists and Aquents occur extensively above limestone in the Everglades and the soil with the second highest As concentration (38.2 mg kg-1) was a shallow Aquept overlying limestone in a wet forest on the Gulf coast. It is possible that the high As levels in these soils were accumulated via biological processes associated with long-term inundation or saturation of the soils. Another possibility could be the release of As contained in underlying bedrock (limestone or phosphate deposits), since these sedimentary rocks contain relatively high As concentrations (Smith et al., 1998). A comparison between As concentrations in surface marl soil and bedrock (limestone) of the ENP samples indicated that As originated from the bedrock rather than from anthropogenic sources (Chen et al., 2000). In addition, as a group these wet, As-rich soils had significantly ({alpha} < 0.05) higher contents of organic C, clay, and total Fe, Al, and P, as compared with those of the remaining soils (Table 4). Leaching of As from wetland soils could be inhibited by clay, calcite, and organic matter, as well as by the hydrated oxides of Fe and Al, with As (III) being more thermodynamically stable than As (V) (Smith et al., 1998). The flooding and drainage cycles in wetland soils favor the conversion of Fe mineralogy from a more crystalline to a less crystalline phase, which will increase the surface area and number of potential As sorption sites (McGeehan et al., 1998).


View this table:
[in this window]
[in a new window]
  		Table 4. Comparison of baseline soil-As concentration and selected properties of upland and wetland soils.

 
It should be noted, however, that not all of the Aquic (wet) suborders were high in As. The most notable exception was the Aquods, which contained the lowest As concentration (Table 1). Aquods tend to have a fluctuating water table and are generally not saturated at the surface for long periods. In effect, they are wet (poorly drained) but normally not wet enough (very poorly drained) to be considered wetland soils. They are most often found in the flatwood landscapes that constitute the local uplands in Florida. Aquods are also characterized by the fact that they have undergone podzolization, which promotes near-surface depletion of Fe and Al leading to a reduction in As retention capacity.


   CONCLUSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
REFERENCES
 
This study showed that background concentrations of As in Florida surface soils varied greatly based on soil type and properties. Multiple regression analysis indicated concentrations of total Fe and P are two most important soil properties that influence soil As concentration. Soils developed in wetland environments (Saprists, Hemists, Aquents, Aquepts, and Aquolls) tended to be naturally high in As concentration compared with soils developed in upland environments (Aquods, Humods, Psamments, Udalfs, and Udults). Sources of As in Florida wetland soils may result from the weathering of underlying bedrock (limestone or phosphate deposition), or from bioaccumulation of As by aquatic organisms. Caution should be taken when a single soil As concentration (UBC) is used as a soil cleanup standard or for regulatory purposes.


   ACKNOWLEDGMENTS
 
This research was sponsored in part by the Florida Center for Solid and Hazardous Waste Management (Contract No. 96011017). The authors thank Drs. Y.C. Li and G.W. Hurt for generously providing marl soil samples from south Florida. We are also indebted to those who participated in the Florida Cooperative Soil Survey. Their collection and characterization of a large number of Florida soil samples made this study possible. The helpful comments and suggestions made by Drs. R. Brown, J. Thomas, and three anonymous reviewers are gratefully acknowledged.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 
Approved for publication as Florida Agricultural Experiment Station Journal Series No. R-07010.

Received for publication September 15, 2000.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 ELEMENTSHome page
G. Morin and G. Calas
Arsenic in Soils, Mine Tailings, and Former Industrial Sites
Elements, April 1, 2006; 2(2): 97 - 101.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Chen, M.
Right arrow Articles by Harris, W. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Chen, M.
Right arrow Articles by Harris, W. G.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Chen, M.
Right arrow Articles by Harris, W. G.
Related Collections
Right arrow Other Soil Types
Right arrow Heavy Metals

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		 Journal of
Environmental Quality