Water Air Soil Pollut (2009) 204:385–398 DOI 10.1007/s11270-009-0052-6
Arsenic Risk Assessment: The Importance of Speciation in Different Hydrologic Systems C. T. Markley & B. E. Herbert
Received: 29 November 2008 / Accepted: 16 March 2009 / Published online: 7 April 2009 # Springer Science + Business Media B.V. 2009
Abstract The processes impacting arsenic toxicity are a function of molecular speciation, where risk from chronic exposure to the reduced arsenic species is estimated to be four orders of magnitude higher than many oxidized arsenic species. While the adverse health effects of arsenic are generally well known, the impact of speciation on carcinogenic and noncarcinogenic adverse health effects has rarely, if ever, been considered in traditional chronic arsenic exposure risk assessments. Utilizing standard Environmental Protection Agency protocol, lifetime cancer risk and hazard quotient are calculated for chronic arsenic exposure at the local, regional, and national scale to characterize potential risk as a function of arsenic speciation. Additionally, the antagonistic and synergistic impacts of biogeochemical processes on arsenic bioavailability and bioaccessibility are discussed and show chronic exposure risk is likely to be
C. T. Markley (*) : B. E. Herbert Department of Geology and Geophysics, Texas A&M University, College Station, TX 77843-3115, USA e-mail:
[email protected] C. T. Markley U.S. Nuclear Regulatory Commission, MS: EBB 2-BO2, Washington, DC 20555-0001, USA
reduced below some maximum value calculated for reduced arsenic species. Keywords Arsenic . Speciation . Risk assessment
1 Introduction The adverse health effects of arsenic are well known, where chronic exposure can impact the respiratory, gastrointestinal, cardiovascular, nervous, and hematopoietic systems (Jain and Ali 2000). Chronic arsenic exposure is manifested in conditions such as skin lesions, carcinoma, keratosis, and blackfoot disease (Mandal et al. 1998). Further, the Environmental Protection Agency (EPA) categorizes arsenic as a Class A human carcinogen. Arsenic speciation greatly impacts the processes that result in arsenic toxicity. The oxidized species, arsenate (As5+), is a molecular analog of phosphate and can inhibit oxidative phosphorylation (Oremland and Stoltz 2003). The reduced species, arsenite (As3+), is considered a more toxic arsenic species because of its ability to bind with sulfhydryl groups, thereby impacting the function of a broad range of proteins and enzymes (National Research Council 1999). Pentavalent organoarsenicals are generally considered the least toxic arsenic species and include monomethylarsonic acid (MMA5+), dimethylarsinic acid (DMA5+), and numerous higher molecular weight arsenosugars (Jain and Ali 2000;
386
Oremland and Stoltz 2003). In a long-term study of ingestion of various arsenicals by rats, Yoshida et al. (1998) showed that DMA5+, arsenobetaine, and, to a lesser extent, MMA5+, were excreted in urine relatively unchanged suggesting minimal intracellular uptake and therefore lower toxicity when compared to inorganic arsenic species. Conversely, trivalent organoarsenicals have been shown to be an order of magnitude more toxic than arsenite in Chang Human Hepatocytes (Petrick et al. 2000). Though there is an increased understanding with respect to the mechanisms of arsenic toxicity, arsenic speciation has yet to be incorporated into risk assessments associated with chronic exposure. The arsenic cycle is influenced by numerous biological and geochemical processes which control arsenic speciation, distribution, and potential toxicity. Arsenic fate and transport in oxic surface water systems are generally considered a function of sorption/desorption and precipitation/dissolution reactions with hydroxide minerals (La Force et al. 2000; Smedley and Kinniburgh 2002) while precipitation/dissolution reactions with sulfide minerals dominate arsenic fate and transport in anoxic systems (Smedley and Kinniburgh 2002). Microorganisms, including bacteria and phytoplankton, exert an influence on arsenic speciation through a variety of detoxification and/or respiratory processes. In oxic surface waters, phytoplankton have been shown to reduce arsenate to arsenite in the log phase of growth and biotransform arsenic to the various organic species when growth had reached steady state, thereby initially increasing potential toxicity and subsequently reducing potential toxicity based on the resultant arsenic species (Andreae 1979; Hasegawa et al. 2001). Further, many microbes have the ability to oxidize arsenite, as either a detoxification mechanism or as an electron donor (Oremland and Stoltz 2003). Indirectly, microbes alter arsenic speciation via processes impacting the redox conditions of the system. For example, seasonal anoxia in lacustrine systems caused by organic matter decomposition could result in arsenite dominance. The complex biogeochemical arsenic cycle demonstrates the potential for arsenic speciation to be dominated by species not predicted based on thermodynamics alone. Often, risk assessments of arsenic
Water Air Soil Pollut (2009) 204:385–398
have been performed based on total concentrations or bioavailability of arsenic and neglected the impact of speciation (Gupta et al. 1996; Rapant and Krčmová 2007). Based on the various mechanisms that lead to arsenic toxicity, arsenic speciation is expected to greatly influence potential risk. Because of this, carcinogenic and noncarcinogenic risk was estimated at the national, regional, and local scale based on arsenic concentration and speciation and utilized standard EPA protocol.
2 Materials and Methods 2.1 Development of the Chronic Exposure Risk Charts The Environmental Protection Agency defines risk as the probability of some adverse health effect that results from exposure to some contaminant in the environment. Risk assessments characterize carcinogenic and noncarcinogenic adverse effects by calculating the lifetime cancer risk and hazard quotient, respectively, based on exposure concentration, duration, and pathways. For chronic residential exposure, the EPA recommends an acceptable carcinogenic risk level of 10−6, where exposure would result in a 1 in 1,000,000 chance of developing cancer. For industrial purposes, the recommended carcinogenic risk level is 10−4. These risk levels are generally conservative when compared to the likelihood of developing cancer. The American Cancer Society (2007) reported that the probability of developing an invasive cancer at some point in life was 37.9% for females and 45.3% for males in the USA for the time period 2001–2003. Four chronic exposure risk charts (CERCs) were developed using standard EPA protocol to calculate arsenic risk based on arsenic concentration and speciation. These CERCs visualize lifetime cancer risk and hazard quotient for both sediment and water ingestion pathways based on adult exposure. Chronic daily intake (CDI; milligrams per kilogram per day) was determined via the equation: CDI ¼ CS IR EF CF1 1=BW 1=AT
ð1Þ
where CS was arsenic concentration (milligrams per kilogram, milligrams per liter), IR was the
Water Air Soil Pollut (2009) 204:385–398
387
ingestion rate (milligrams per day, liter per day), EF was the exposure frequency (day), CF1 was a conversion factor, BW was body weight (kilograms) and AT was the average time of exposure (year). The noncarcinogenic hazard quotient (HQ) and lifetime cancer risk (LCR) were calculated via the equations: HQ ¼ CDI=RfD
ð2Þ
RF ¼ CDI OCSF
ð3Þ
where RfD was the reference dose (milligrams per kilogram per day) and OCSF was the oral cancer slope factor. Parameters used in calculating the lifetime cancer risks and hazard quotients were based on standard EPA assumed values (Table 1). Limited data exist for speciation-based arsenic reference doses (RfD) and were therefore estimated based on the relative toxicity between the species (Table 2). Because of the limited data on the Table 1 Parameters used to calculate arsenic risk
Parameter
Units
Value
Description
Source
LCR
–
Calculated
Calculated lifetime cancer risk
b
HQ
–
Calculated
Calculated hazard quotient
c
IRs
mg day−1
100
Ingestion rate of soil
EPA
b
LCR=CDI×OCSF
c
HQ=CDI/RfD
d
CDI = CS × IR × EF × ED × CF1×1/BW×1/AT
−1
IRl
l day
2
Ingestion rate of liquid
EPA
EF
day year−1
350
Exposure frequency
EPA
ED
year
24
Exposure duration
EPA
CF1
kg mg−1
0.000001
Conversion factor
–
BW
kg mg−1
70
Body weight
EPA
AT-C
day
25,550
Average time (cancer)
EPA
CDIc
mg kg−1 day−1
Calculated
Chronic daily intake (cancer)
d
AT-NC
day
8,760
Average time (noncancer)
EPA
CDIn
mg kg−1 day−1
Calculated
Chronic daily intake (noncancer)
d
OCSF
–
1.5
Oral cancer slope factora
EPA
a
Estimated based on multiple published studies
impact of chronic arsenic exposure to toxic effects, the values observed in acute toxicity experiments are assumed to be proportional to chronic effects. Acute As3+ toxicity was observed at concentrations ranging from approximately two to 100 times less than As5+ (Salisbury 2006, Teruaki and Kitao 2003) while acute MMA 5+ and DMA 5+ toxicity was observed at concentrations approximately three to 3,000 times more than As5+ (Kuroiwa et al. 1995; Hirano et al. 2004). Acute MMA3+ toxicity consistently occurred at concentrations lower than As3+. However, MMA3+ is not considered here because trivalent methylated arsenic species are expected to oxidize to MMA5+ in natural systems within days significantly reducing potential exposure (Gong et al. 2001). The Integrated Risk Information System (IRIS) provides an oral arsenic RfD based on studies showing the relationship between blackfoot disease, age, and dose (Tseng 1977). However, the IRIS does not provide RfDs for the individual arsenic species. Subsequent studies in the Blackfoot Disease Area (Taiwan) showed As3+ and As5+ proportions approximately 69% and 26%, respectively (Chen et al.
RfD
–
1.5
Oral cancer slope factor [As ]
–
75
Oral cancer slope factor [As3+]
est.
–
0.015
Oral cancer slope factor DMA5+
est.
mg kg−1 day−1
0.0003
Oral reference dose valuea
EPA
mg kg−1 day−1
0.0003
Oral reference dose value [As5+]
mg kg
−1
day
−1
mg kg−1 day−1
5+
3+
est.
est.
0.000006
Oral reference dose value [As ]
est.
0.03
Oral reference dose value [DMA5+]
est.
388
Water Air Soil Pollut (2009) 204:385–398
Table 2 Arsenic toxicity as a function of speciation Receptor
Chang human hepatocytes Chang human hepatocytes Chang human hepatocytes CAsE cells
Statistic Units
As5+
As3+
MMA5+
DMA5+
MMA3 DMA3 Notes +
+
References
LC50
μM
–
68
–
–
6
–
–
Petrick et al. (2000)
LC50
μM
–
19.8
–
–
6.3
–
–
Petrick et al. (2000)
LC50
μM
–
164
–
–
13.6
–
–
Petrick et al. (2000)
LC50
μM
140
= DMA5 = MMA5 –
–
–
Romach et al.(2000)
Endothelial cells
LC50
μM
4,000 + –
220
–
–
36
–
–
Hirano et al. (2003)
Endothelial cells
LC50
μM
220
36
36,600
2,540
4.6
–
–
Hirano et al. (2004)
30.3
14.6
–
–
–
–
–
Suhendrayatna et al. (2002)
100
300
–
–
–
–
Organoarsenic Kuroiwa et al. oxidation (1995) state assumed 5+ – He et al. (2008)
−1
+
+
Oryzias latipes
LC50
mg l
Macrobrachium rosenbergii
LC50
μg cm−3
30
–
Daphnia carinata
LC50
mg l−1
1.499
0.554
–
–
Daphnia pulex
LC50
μg l−3
2,300–3,900
13,800–15,700
120
–
Salmo gairdneri
LC50
mg l−1
114.1
17.7
–
–
–
–
Salmo gairdneri
LC50
mg l−1
58
20.7
–
–
–
–
15°C
McGeachy and Dixon (1989)
Rat liver TRL 1215 cells Zebrafish larvae
LC50
μM
–
20
–
1,500
–
–
–
Sakurai (2003)
LC50
μM
1,347
771.98 –
–
–
–
–
Salisbury (2006)
Zebrafish larvae
EC50
μM
1,172
570
–
–
–
–
–
Salisbury (2006)
Macrophages
LC50
MU.M
500
5
–
–
–
–
–
Teruaki and Kitao (2003)
Macrophages
LC50
mM
–
–
>10
5
–
–
Toxicity not observed in MMA5+
Teruaki and Kitao (2003)
1994). To simplify calculations and provide a conservative estimate of risk, the oral RfD listed was assumed to represent chronic arsenate exposure. Based on the ranges in the literature, As3+ was assumed to be 1.5 orders of magnitude more toxic than As5+ and the pentavalent methylated arsenic species were assumed to be two orders of magnitude less toxic than As5+. RfD values were adjusted based on these relative toxicities.
Shaw et al. (2007) Organic and inorganic values are not differentiated 5°C McGeachy and Dixon (1989)
2.2 Calculation of Chronic Exposure Risk The associated LCR and HQ for chronic adult exposure are linearly related to the arsenic species concentration in milligrams per liter and are summarized by the equations: Arsenite LCR ¼ 2:4658 As3þ mg l1 5e15
ð4Þ
Water Air Soil Pollut (2009) 204:385–398
LCR ¼ 0:0411 As5þ mg l1
389
ð6Þ
resource, risk associated with a complex mixture of arsenic species could be assessed. Risk was calculated for each individual sample by utilizing the appropriate equation (Eqs. 4–9) for each arsenic species with the sum representing total arsenic risk of each sample.
ð7Þ
3 Results and Discussion
ð5Þ
Arsenate LCR ¼ 0:0411 As5þ mg l1
HQ ¼ 91:324 As5þ mg l1
DMA5+, MMA5+
3.1 Description of the Chronic Exposure Risk Charts
LCR ¼ 0:0004 XMA5þ mg l1 þ 6e19
ð8Þ
HQ ¼ 0:9132 XMA5þ mg l1 1e15 :
ð9Þ
Arsenic concentrations used to determine nationalscale arsenic risk were a product of the U.S. Geological Survey’s National Water-Quality Assessment Program (Ryker 2001). The groundwater dataset included over 20,000 samples collected from potable groundwater resources and therefore did not include thermal or saline waters. All samples were collected between 1973 and 2001 and were measured via inductively coupled plasma mass spectrometry or hydride generation atomic absorption spectroscopy. Only total arsenic concentrations were reported so to estimate risk at the national scale, chronic exposure risk was calculated for each arsenic species (As 3+ , As 5+ , pentavalent methylated arsenicals) to assess the impact of speciation on potential risk utilizing Eqs. 4–9. Arsenic concentrations used to determine regionalscale arsenic risk were extracted from the National Uranium Reconnaissance Evaluation (NURE) dataset (U.S. Geological Survey 2004). The NURE dataset included total arsenic concentrations collected in the 1970s from the Gulf Coast Uranium Province surface waters. Again, chronic exposure risk was calculated for each arsenic species (As3+, As5+, pentavalent methylated arsenicals) to assess the impact of speciation on potential risk utilizing Eqs. 4–9. Local-scale arsenic concentration data were obtained for Burton Creek (Brazos Co., TX, USA) from Markley and Herbert (2006). Samples were collected at two time periods and included arsenic speciation data. Though not a drinking water
The lifetime cancer risk and hazard quotient associated with chronic arsenic exposure in drinking water can be divided into four categories based on the calculated risk and arsenic concentrations: minimal, low, high, and extreme (Fig. 1). Region I represents minimal risk and occurs where arsenic concentrations and associated risk are below EPA limits, while region IV is extreme risk and occurs where both arsenic concentration and associated risk are above EPA limits. Region III is high risk and occurs where arsenic concentrations are below drinking water standards, but associated risk is above the EPA limit as is possible with chronic As3+ exposure. Region III is not considered extreme risk because the lower arsenic concentrations are more susceptible to factors influencing bioavailability. In other words, sorption and deposition could rapidly decrease associated risk because of the low As concentrations. Region II is low risk and occurs where arsenic concentrations are above drinking water standards, yet risk is below
Fig. 1 General categorization of chronic arsenic exposure risk
390
Water Air Soil Pollut (2009) 204:385–398
EPA-recommended values. Region II risk could be associated with relatively high arsenic concentrations, yet low bioavailability resulting because of various sequestration processes. CERCs showing calculated risk associated with chronic arsenic exposure include the acceptable carcinogenic risk level (<10−6) and noncarcinogenic hazard quotient (<1) as recommended by the EPA (Fig. 2). The EPA and World Health Organization have set the drinking water standard for arsenic at 10 μg l−1. However, the CERCs show that chronic exposure at that level is well above the acceptable carcinogenic risk level for all three arsenic species, with risk of developing cancer from chronic arsenite exposure at greater than 10−2, over four orders of magnitude above the acceptable carcinogenic risk level. Cancer risk may be even greater for chronic
exposure to reduced organoarsenicals based on studies showing higher toxicity than reduced inorganic arsenic (Petrick et al. 2000; Hirano et al. 2003). The impact of arsenic speciation on calculated risk is quite dramatic, with arsenite risk at almost four orders of magnitude higher than the methylated arsenic species. It is important to note that the speciation based CERCs describe the maximum potential risk based on speciation and concentration and characterizes surface water systems where arsenic bioavailability is generally considered 100%. Risk in these systems would plot between As3+ and DMA5+ calculated risk. Conversely, arsenic bioavailability in sediments is highly variable and depends on the geochemical characteristics of the system. For example, Yang et
Fig. 2 Lifetime cancer risk (a) and hazard quotient (b) for ingestion of sediments in adults calculated using standard Environmental Protection Agency protocol. Cancer risk (c) and hazard quotient (d) for ingestion of drinking water in adults.
EPA-LCR and EPA-HQ are the Environmental Protection Agency acceptable levels for lifetime cancer risk (10−6) and hazard quotient (1)
Water Air Soil Pollut (2009) 204:385–398
al. (2002) showed that lower pH and increased iron oxide content greatly reduced arsenic bioavailability through sorption processes that subsequently sequestered arsenic. In other words, arsenic risk would plot below the DMA5+-based exposure risk. Because of the dynamic nature of the arsenic cycle, the above equations present a fast and efficient way to estimate maximum potential risk based on speciation and concentration data. 3.2 Arsenic Speciation in the Environment The biogeochemical arsenic cycle is impacted by numerous differing processes in surface- and ground water systems that allow for predictable trends in arsenic speciation. A summary of arsenic speciation ranges for select aqueous environments is presented in Table 3 and shows inorganic arsenic species dominate groundwater systems while arsenic speciation in surface water systems is much more variable. Redox conditions appeared to control groundwater arsenic speciation, where arsenite dominated anoxic systems and arsenate dominated oxic systems characterized by volcanic rock with neutral to high pH. In surface waters, methylated arsenic species have been shown to dominate the epilimnion and comprise up to 96% of total arsenic while arsenite has been shown to dominate the hypolimnion and comprise up to 83%. The impact of arsenic speciation on potential risk at the national, regional, and local scales is discussed below. 3.3 Application of the Chronic Exposure Risk Charts 3.3.1 National Assessment: Groundwater Groundwater arsenic concentrations in the USA are highly variable, averaging 7.37±48.1 μg l−1 with maximum concentrations measured at 2,600 μg l−1 based on the data in Ryker (2001), though anthropogenic activities can result in arsenic concentrations reaching 100,000 μg l−1 (Nordstrom 2002). Understanding the impact of arsenic speciation in groundwater resources is important because it provides a drinking water source for more than 99% of the rural USA population (Ryker 2001). National-scale arsenic risk was plotted assuming total arsenic concentrations were present as As5+, As3+, or DMA5+ (Fig. 3). Based
391
on the calculations, DMA5+ concentrations would present minimal risk with levels consistently below the EPA-recommended 10−6 acceptable level. Calculating risk assuming arsenic was present as arsenite substantially increased the risk associated with exposure to levels above 10−4. Utilizing speciation data provide a more accurate view of risk from chronic arsenic exposure and therefore can be utilized to better target potential areas of concern. Based on general trends in arsenic speciation in groundwater systems, arsenic risk would likely occur as inorganic species and would therefore plot between the As5+ and As3+ calculated risks on the CERCs. 3.3.2 Regional Assessment: Surface Water Lake Corpus Christi is a surface water reservoir in South Texas that provides drinking water for approximately 350,000 people in the Corpus Christi metropolitan area. Arsenic concentrations in the region are derived from weathering of volcanic ash and have been measured up to 44 μg l−1 (U.S. Department of Energy 1995). To determine the impact of arsenic speciation, regional scale arsenic risk was plotted using data extracted from the NURE dataset (U.S. Geological Survey 2004) and assumed total arsenic concentrations were present as As5+, As3+, or DMA5+ (Fig. 4). The impact of arsenic speciation was instantly clear, where arsenic present as As3+ resulted in lifetime cancer risk above the EPA-recommended level, while As5+ and DMA5+ generally remained below the acceptable risk level. Speciation in this reservoir is likely a mixture of As5+, DMA5+, and, to a lesser extent, As3+. Sohrin et al. (1997) showed that arsenic speciation in surface waters is seasonal and resulted from biologically mediated processes that reduced arsenate to arsenite and methylated arsenic to create a temporally dynamic mixture of arsenic species. 3.3.3 Local Assessment Arsenic concentrations in the contaminated stream ranged from approximately 16 to 90 μg l−1 and were dominated by the DMA species (Fig. 5). The DMA is assumed to be the pentavalent species due to the rapid oxidation of trivalent organoarsenicals in natural systems (Gong et al. 2001). Although the arsenic concentrations were at least 78% DMA and indicated
392
Water Air Soil Pollut (2009) 204:385–398
Table 3 Arsenic speciation in select surface and groundwaters Location
Water
Comments
Units
As5+
As3+
Gironde, Biscay Bay
Estuary
Epilimnion
%
57.9–65.1 14.5–20.7
0.0
9.6– – 11.7
Michel et al. (1997)
Estuary
Hypolimnion
%
63.5–99.0 1–18.2
0–0.5
0–8.5 –
Michel et al. (1997)
Tinto River
Estuary
Acidic
%
<0.01– 99.1
<0.01–96.5
nd
nd
–
Elbaz-Poulichet et al. (2000)
Twin Butte Vista
Hot spring
Alkaline
%
24–76
24.4–76
nd
nd
–
Gihring et al. (2001)
Pavin crator lake
Freshwater lake (meromictic)
Mixolimnion
%
88–93
5.5–12
nd
nd
–
Seyler and Marting (1989)
monimilimnion
%
10–48.9
12–90
nd
nd
–
Seyler and Marting (1989)
Up to – 64
MMA DMA OrgAs References
Lake Biwa
Freshwater lake
Epilimnion
%
–
Up to 94
–
Davis Creek Reservoir
Freshwater lake
Epilimnion
%
29–47
0.39–11
11–18 23– 54
–
Anderson and Bruland (1991)
Freshwater Lake
Hypolimnion
%
1.0–76
27–83
9.7– 14
1.3– 28
–
Anderson and Bruland (1991)
Freshwater lake
Epilimnion
%
28–62
30–48
–
–
–
Aurillo et al. (1994)
Freshwater lake
Hypolimnion
%
Up to 78
–
–
–
–
Aurillo et al. (1994)
Burton Creek, TX
Creek
Ephemeral
%
0.46–2.5
1.7–47
1.0– 5.0
46– 95
–
Markley and Herbert (2006)
North Haiwee Reservoir
Freshwater lake
%
73.7–89.4 10.6–26.3
nd
nd
–
Kneebone et al. (2002)
Freshwater Lake, Maine
Pore water
Landfill impacted
%
0–12
88–100
nd
nd
–
Nikolaidis et al. (2004)
Moira Lake, Ontario
Surface water
Mining impacted
μg l-1
18.8–58
7.4–74.7
nd
nd
0.01– 1.1
Azcue and Nriagu (1995)
μg l-1
0–110
0–196
nd
nd
0–10
Azcue and Nriagu (1995)
–
Mainly as As5+
Del Razo et al. (1990)
–
Mainly as As5+
Nicolli et al. (1989)
Upper Mystic Lake
Pore water Mexico (Lagunera)
Argentina (ChacoPampean Plain)
Tulare Basine
Groundwater
Groundwater
Oxic, nuetral to high pH Holocene loess/ rhyolitic ash Oxic, neutral– high pH,
Groundwater
San Joaquin Valley, CA
Lower Illinois River Basin
Volcanic seds
Groundwater
Saline, high alkalinity Holocene, basin fill sediments Internally drained, mixed redox High salinity in some shallow Glacial drift aquifer
Sohrin et al. (1997)
Smedley et al. (2002)
–
%
As3+ increases with depth
0–47
50–108
Fujii and Swain (1995)
nd
nd
–
Warner (2001)
Water Air Soil Pollut (2009) 204:385–398
393
Table 3 (continued) Location
Water
Comments
Units
As5+
As3+
MMA DMA OrgAs References
Northern-tier PA Counties
Groundwater
Unconsolidated glacial deposits
%
1.4–120
6.5–98.8
bdl– 1.9
– irrelevant or not reported, nd not determined, bdl below detection limit Fig. 3 Hypothetical chronic exposure risk in USA groundwater based on total arsenic concentration data from Ryker (2001). Chronic exposure risk was calculated for three hypothetical scenarios: arsenic concentrations present as either As5+, As3+, or the oxidized methylated species. Calculations assume adult exposure via oral ingestion of drinking water
bdl– 3.2
–
Low and Galeone (2007)
394
Water Air Soil Pollut (2009) 204:385–398
Fig. 4 Hypothetical chronic exposure risk in surface water based on total arsenic concentration data from the U.S. Geological Survey (2004). Chronic exposure risk was calculated for three hypothetical scenarios: arsenic concentrations present as either As5+, As3+, or the oxidized methylated species. Calculations assume adult exposure via oral ingestion of drinking water
arsenic in the surface water was highly bioavailable, risk associated with DMA was minimal. Conversely, arsenite represented a minor fraction of total arsenic in this system, yet accounted for nearly 100% of the associated risk demonstrating the importance of utilizing speciation data for arsenic risk assessments. The inset shows that arsenic at these concentrations are above the drinking water standard for arsenic and the permissible risk level and are therefore considered to be an extreme risk (region IV). Fortunately, this is not a drinking water resource, meaning ecological impacts are the primary concern. However, this clearly highlights the importance of utilizing speciation data when assessing risk associated with chronic arsenic exposure. 3.4 The Role of Bioavailability and Bioaccessibility Total arsenic concentrations do not indicate true potential risk from exposure and often overestimates the associated risk because arsenic may be sequestered in nonbioavailable forms. Gupta et al. (1996) described a three-level risk assessment that incorporated arsenic bioavailability and bioaccessibility to develop economic risk management strategies. Bioavailable arsenic is defined as the mobilized fraction of arsenic in pore and surface water systems. Bioaccessible arsenic is generally considered immobilized through sequestration processes such as sorption to hydroxide mineral surfaces, yet has the
potential to become bioavailable. Numerous geochemical factors influence arsenic bioavailability and bioaccessibility. Arsenic desorption via ion exchange may mobilize arsenic and therefore increase bioavailability. Further, arsenic coprecipitation with minerals such as oxides or sulfides may create a nonbioavailable fraction. These sorption/desorption and precipitation/dissolution reactions play a major role affecting arsenic speciation and distribution. For example, Dixit and Hering (2003) found that As5+ sorption to iron hydroxides and goethite is more favorable at pH<5–6 while As3+ sorption to these minerals is more favorable at pH>7–8. At lower pH values, As3+ would be more mobile, increasing bioavailability and potential risk associated with chronic exposure. Fendorf et al. (2004) showed that arsenic bioaccessibility decreased over time when spiked into uncontaminated A and B soil horizons collected from Melton Valley soil and incubated for 400 days. Extractable arsenic concentrations decreased during the first 30 days of the experiment, likely due to high concentrations of Fe oxides (10.68–22.07 g kg−1) sequestering arsenic. While this indicated iron oxides effectively scavenged bioavailable arsenic, the study was limited in that the system was unperturbed through the duration of the experiment. For example, the introduction of competing anions such as phosphate could effectively replace a portion of arsenic on the mineral surface, thus remobilizing the arsenic (Hongshao and Stanforth 2001). Also, the development of anoxic
Water Air Soil Pollut (2009) 204:385–398
395
Fig. 5 Arsenic speciation (a), chronic exposure risk (b), and hazard quotient (c) associated with arsenic in a contaminated stream. The inset shows where these samples would plot on the appropriate CERC. Chronic exposure risk was calculated using the above arsenic concentration and speciation data. Calculations assume adult exposure via oral ingestion of drinking water
conditions could trigger arsenic release through the reductive dissolution of these oxide minerals, thereby biasing the results towards decreased bioavailable values (Smedley and Kinniburgh 2002). In turn, these values would underestimate bioavailable and bioaccessible based risk assessments. Determining potential arsenic risk based on speciation data provides a more accurate representation of potential lifetime cancer risk than utilizing arsenic
data based on bioaccessibility. Using a standard reference soil (SRM 2710) from the National Institute of Standards and Technology, Ellickson et al. (2001) determined arsenic bioaccessibility and oral bioavailability values of approximately 66% and 38%, respectively. This is less than a one order of magnitude deviation from total concentration, much less than the four order of magnitude difference in arsenic toxicity based on speciation shown above.
396
3.5 Assessment of the Chronic Exposure Risk Chart and Potential Limitations Inherent limitations are involved when performing any risk assessments and/or epidemiological studies. Different populations may have a genetic predisposition that increases susceptibility to the adverse health effects of chronic arsenic exposure. In one example, Wang et al. (2003) reported an increased prevalence for females to develop diabetes in the arseniasisendemic area in Taiwan. There are also differences in lifestyle that can impact the accuracy of risk assessments such as poverty level or exposure via food. Based on the CERC, arsenite concentrations would have to be lower than 1 ng l−1 to reduce the risk below the EPA-recommended 10−6 acceptable level. However, the cost to reduce arsenic to concentrations below this level would be prohibitive. An analysis performed to assess the cost of reducing arsenic concentrations in drinking water estimated that it would cost approximately $693 million dollars to lower arsenic concentrations to 3 μg l−1 (Stedge 2000). Further, the benefit of lowering arsenic concentrations below these levels would be minimal when compared to the lifetime probability of females (37.9%) and males (45.3%) of developing some form of invasive cancer (American Cancer Society 2007). These limitations do not diminish the ability of the CERCs to help identify potential areas of concern in baseline risk assessments by utilizing arsenic speciation data to determine areas of increased risk.
4 Implications It is evident that arsenic speciation greatly impacts potential risk associated with chronic arsenic exposure. Arsenic bioavailability is approximately 100% when exposed via ingestion of drinking water and therefore is accurately represented on the arsenic species-based risk plot. Bioavailability based on exposure via ingestion of arsenic-contaminated sediments will be highly dependent on the geochemical associations such as the presence of iron oxides and therefore provide a higher range of toxicity. Regardless of exposure pathway, this study illustrates the need for individual RfD values for the major arsenic species: As3+, As5+, and the pentavalent methylated species. Further, the development of species specific
Water Air Soil Pollut (2009) 204:385–398
RfD values creates another tool that can be used to identify regions at higher risk from chronic arsenic exposure. Disclaimer One of the authors of this paper is an employee of the U.S. Nuclear Regulatory Commission (NRC). This work was completed while both authors were affiliated with Texas A&M University. Final revisions were completed during nonworking hours while affiliated with the NRC.
References American Cancer Society. (2007). Cancer Facts & Figures 2007. Atlanta, GA: American Cancer Society. http://www. cancer.org/ downloads/STT/CAFF2007PWsecured.pdf Anderson, L. C. D., & Bruland, K. W. (1991). Biogeochemistry of arsenic in natural waters: The importance of methylated species. Environmental Science & Technology, 25(3), 420– 427. doi:10.1021/es00015a007. Andreae, M. O. (1979). Arsenic speciation in seawater and interstitial waters: The influence of biological–chemical interactions on the chemistry of a trace element. Limnology and Oceanography, 24(3), 440–452. Aurillo, A. C., Mason, R. P., & Hemond, H. F. (1994). Speciation and fate of arsenic in three lakes of the Aberjona Watershed. Environmental Science & Technology, 28, 577–585. doi:10.1021/es00053a008. Azcue, J. M., & Nriagu, J. O. (1995). Impact of abandoned mine tailings on the arsenic concentrations in Moira Lake, Ontario. Journal of Geochemical Exploration, 52, 81–89. doi:10.1016/0375-6742(94)00032-7. Chen, S.-L., Dzeng, S. R., Yang, M.-H., Chiu, K. H., Shleh, G.-M., & Wai, C. M. (1994). Arsenic species in groundwaters of the Blackfoot Disease Area, Taiwan. Environmental Science & Technology, 28, 877–881. doi:10.1021/ es00054a019. Del Razo, L. M., Arellano, M. A., & Cebrián, M. E. (1990). The oxidation states of arsenic inwell-water from a chronic arsenicism area of northern Mexico. Environmental Pollution, 64, 143–153. doi:10.1016/0269-7491 (90)90111-O. Dixit, S., & Hering, J. G. (2003). Comparison of arsenic(V) and arsenic(III) sorption onto iron oxide minerals: implications for arsenic mobility. Environmental Science & Technology, 37, 4182–4189. doi:10.1021/es030309t. Elbaz-Poulichet, F., Dupuy, C., Cruzado, Z. V., Achterberg, E. P., & Braungardt, B. (2000). Influence of sorption processes by iron oxides and algae fixation on arsenic and phosphate cycle in an acidic estuary (Tinto River, Spain). Water Research, 34(12), 3222–3230. doi:10.1016/S00431354(00)00073-7. Ellickson, K. M., Meeker, R. J., Gallo, M. A., Buckley, B. T., & Lioy, P. J. (2001). Oral bioavailability of lead and arsenic from a NIST standard reference soil material. Archives of Environmental Contamination and Toxicology, 40, 128– 135. doi:10.1007/s002440010155. Fendorf, S., La Force, M. J., & Guangchao, L. (2004). Heavy metals in the environment: Temporal changes in soil
Water Air Soil Pollut (2009) 204:385–398 partitioning and bioaccessibility of arsenic, chromium and lead. Journal of Environmental Quality, 33, 2049–2055. Fujii, R., & Swain, W. C. (1995). Areal distribution of selected trace elements, salinity and major ions in shallow ground water, Tulare Basin, Southern San Joaquin Valley, California. US. U.S. Geologic Survey Water Resources Investigation. Sacramento, CA: USGS. Gihring, T. M., Druschel, G. K., McCleske, R. B., Hamers, R. J., & Banfield, J. F. (2001). Rapid arsenite oxidation by Thermus aquaticus and Thermus thermophilus: field and laboratory investigations. Environmental Science & Technology, 35, 3857–3862. Gong, Z., Lu, X., Cullen, W. R., & Le, C. L. (2001). Unstable trivalent arsenic metabolites, monomethylarsonous acid and dimethylarsinous acid. Journal of Analytical Atomic Spectrometry, 16, 1409–1413. doi:10.1039/b105834g. Gupta, S. K., Vollmer, M. K., & Krebs, R. (1996). The importance of mobile, mobilisable and pseudo total heavy metal fractions in soil forthree-level risk assessment and risk management. Science of the Total Environment, 178, 11–20. doi:10.1016/0048-9697(95)04792-1. Hasegawa, H., Sohrin, Y., Seki, K., Sato, K., Norisuye, K., Naito, K., et al. (2001). Biosynthesis and release of methylarsenic compounds during the growth of freshwater algae. Chemosphere, 43, 265–272. doi:10.1016/S00456535(00)00137-5. He, W., Megharaj, M., & Naidu, R. (2008). Toxicity oftriandpenta-valent arsenic, alone and in combination, to the cladoceran Daphnia carinata: the influence of microbial transformation in natural waters. Environmental Geochemistry and Health, Electronic Publication, http://www.ncbi. nlm.nih.gov/pubmed/19101806. Hirano, S., Cui, X., Li, S., Kanno, S., Kobayashi, Y., Hayakawn, T., et al. (2003). Difference in uptake and toxicity of trivalent and pentavalent inorganic arsenic in rat heart microvessel endothelial cells. Archives of Toxicology, 77(6), 305–312. Hirano, S., Kobayashi, Y., Cui, X., Hayakawa, T., & Shraim, A. (2004). The accumulation and toxicity of methylated arsenicals in endothelial cells: important roles of thiol compounds. Toxicology and Applied Pharmacology, 198 (3), 458–467. doi:10.1016/j.taap. 2003.10.023. Hongshao, Z., & Stanforth, R. (2001). Competitive adsorption of phosphate and arsenate on goethite. Environmental Science & Technology, 35, 4753–4757. doi:10.1021/ es010890y. Jain, C. K., & Ali, I. (2000). Arsenic: occurrence, toxicity and speciation techniques. Water Research, 34, 4304–4312. doi:10.1016/S0043-1354(00)00182-2. Kneebone, P. E., O'Day, P. A., Jone, N., & Hering, J. G. (2002). Deposition and fate of arsenic iniron- andarsenic-enriched reservoir sediments. Environmental Science & Technology, 36, 381–386. doi:10.1021/es010922h. Kuroiwa, T., Yoshihiko, I., Ohki, A., Naka, K., & Maeda, S. (1995). Tolerance, bioaccumulation and biotransformation of arsenic in freshwater prawn (Macrobrachium rosenbergii). Applied Organometallic Chemistry, 9(7), 517– 523. doi:10.1002/aoc.590090704. La Force, M. J., Hansel, C. M., & Fendorf, S. (2000). Arsenic speciation, seasonal transformations andco-distribution with iron in a minewaste-influenced palustrine emergent
397 wetland. Environmental Science & Technology, 34, 3937– 3943. doi:10.1021/es0010150. Low, D. J., & Galeone, D. G. (2007). Reconnaissance of arsenic concentrations in groundwater from bedrock and unconsolidated aquifers in eightnorthern-tier counties of Pennsylvania. Reston, VA: U.S. Department of the Interior, U.S. Geologic Survey. Mandal, B. K., Chowdhury, T. R., Samanta, G., Mukherjee, D. P., Chanda, C. R., Saha, K. C., et al. (1998). Impact of safe water for drinking and cooking on fivearsenicaffected families for 2 years in West Bengal, India. Science of the Total Environment, 218, 185–201. doi:10.1016/S0048-9697(98)00220-4. Markley, C. T., & Herbert, B. E. (2006). Complexity of arsenic biogeochemistry in surface water systems as influenced by a hydrologic event. AGU Fall Meeting. San Francisco, CA: American Geophysical Union. McGeachy, S. M., & Dixon, D. G. (1989). Impact of temperature on the acute toxicity of arsenate and arsenite to rainbow trout (Salmo gairdneri). Ecotoxicology and Environmental Safety, 17(1), 86–93. doi:10.1016/0147-6513(89)90012-2. Michel, P., Boutier, B., Herbland, A., Averty, B., Artigas, L. F., Auger, D., et al. (1997). Behaviour of arsenic in the continental shelf off the Gironde estuary: role of phytoplankton in vertical fluxes during spring bloom conditions. Oceanologica Acta, 21(2), 325–333. doi:10.1016/S03991784(98)80019-4. National Research Council. (1999). Arsenic in drinking water. Washington, DC: National Academy. Nicolli, H. B., Suriano, J. M., Peral, M. A. G., Ferpozzi, L. H., & Baleani, O. A. (1989). Groundwater contamination with arsenic and othertrace-elements in an area of the Pampa, province of Córdoba, Argentina. Environmental Geology and Water Sciences, 14, 3–16. doi:10.1007/BF01740581. Nikolaidis, N. P., Dobbs, G. M., Chen, C., & Lackovic, J. A. (2004). Arsenic mobility in contaminated lake sediments. Environmental Pollution, 129, 479–487. doi:10.1016/j. envpol.2003.11.005. Nordstrom, D. K. (2002). Worldwide occurrences of arsenic in groundwater. Science, 296, 2143–2144. doi:10.1126/ science.1072375. Oremland, R. S., & Stoltz, J. F. (2003). The ecology of arsenic. Science, 300, 939–944. doi:10.1126/science.1081903. Petrick, J. S.,Ayala-Fierro, F., Cullen, W. R., Carter, D. E., & Aposhian, V. (2000). Monomethylarsonous Acid (MMA3+) is more toxic than arsenite in Chang Human Hepatocytes. Toxicology and Applied Pharmacology, 163(2), 203–207. doi:10.1006/taap. 1999.8872. Rapant, S., & Krčmová, K. (2007). Health risk assessment maps for arsenic groundwater content: application of national goechemical databases. Environmental Geochemistry and Health, 29, 131–141. doi:10.1007/s10653-0069072-y. Romach, E. H., Zhao, C. Q., Del Razo, L. M., Cebrián, M. E., & Waalkes, M. P. (2000). Studies on the mechanisms ofarsenic-induced self tolerance developed in liver epithelial cells through continuouslow-level arsenite exposure. Toxicological Sciences, 54, 500–508. doi:10.1093/toxsci/ 54.2.500. Ryker, S. J. (2001). Mapping arsenic in groundwater. Geotimes, 46(11), 34–36.
398 Sakurai, T. (2003). Biomethylation of arsenic is essentially detoxifying event. Journal of Health Science, 49(3), 171– 178. doi:10.1248/jhs.49.171. Salisbury, H.M. (2006). Evaluation of a transgenic zebrafish model for assessing arsenic toxicity. Master of Science thesis, Department of Toxicology, University of Saskatchewan. Seyler, P., & Marting, J.-M. (1989). Biogeochemical processes affecting arsenic species distribution in a permanently stratified lake. Environmental Science & Technology, 23, 1258–1263. doi:10.1021/es00068a012. Shaw, J. R., Glaholt, S. P., Greenberg, N. S.,Sierra-Alvarez, R., & Folt, C. L. (2007). Acute toxicity of arsenic to Daphnia pulex: Influence of organic functional groups and oxidation state. Environmental Toxicology and Chemistry, 26 (7), 1532–1537. doi:10.1897/06-389R.1. Smedley, P. L., & Kinniburgh, D. G. (2002). The review of source, behaviour and distribution of arsenic in natural waters. Applied Geochemistry, 17, 517–568. doi:10.1016/ S0883-2927(02)00018-5. Smedley, P. L., Nicolli, H. B., Macdonald, D. M. J., Barros, A. J., & Tullio, J. O. (2002). Hydrogeochemistry of arsenic and other inorganic constituents in groundwaters from La Pampa, Argentina. Applied Geochemistry, 17, 259–284. doi:10.1016/S0883-2927(01)00082-8. Sohrin, Y., Matsui, M., Kawashima, M., Hojo, M., & Hasegawa, H. (1997). Arsenic biogeochemistry affected by eutrophication in Lake Biwa, Japan. Environmental Science & Technology, 31, 2712–2720. doi:10.1021/ es960846w. Stedge, G. D. (2000). Arsenic in drinking water rule economic analysis. Bethesda, MD: Abt Associates Inc. Suhendrayatna,, Ohki, A., Nakajima, T., & Maeda, S. (2002). Studies on the accumulation and transformation of arsenic in freshwater organisms I. Accumulation, transformation and toxicity of arsenic compounds on the Japanese
Water Air Soil Pollut (2009) 204:385–398 Medaka, Oryzias latipes. Chemosphere, 46(2), 319–324. doi:10.1016/S0045-6535(01)00084-4. Teruaki, S., & Kitao, K. (2003). Nutrition and toxicity of trace elements in immune function systemsarsenic-induced apoptosis in macrophages. Biomedical Research on Trace Elements, 14(4), 259–263. Tseng, W. P. (1977). Effects anddose-response relationships of skin cancer and blackfoot disease with arsenic. Environmental Health Perspectives, 19, 109–119. doi:10.2307/3428460. U.S. Department of Energy. (1995). Baseline risk assessment of ground water contamination at the uranium mill tailings site near Falls City, Texas. Albuquerque, NM: United States Department of Energy. Geological Survey, U. S. (2004). National Uranium Resource Evaluation (NURE) hydrogeochemical and stream sediment reconnaissance data. Denver, CO: U.S. Geological Survey. Wang, S. L., Chiou, J. M., Chen, C. J., Tseng, C. H., Chou, W. L., Wang, C. C., et al. (2003). Prevalence ofnon-insulindependent diabetes mellitus and related vascular diseases in southwesternarseniasis-endemic and nonendemic areas in Taiwan. Environmental Health Perspectives, 3(2), 155– 159. Warner, K. L. (2001). Arsenic in glacial drift aquifers and the implication for drinking water—Lower Illinois River Basin. Ground Water, 39(3), 433–442. doi:10.1111/ j.1745-6584.2001.tb02327.x. Yang, J., Barnett, M. O., Jardine, P. M., Basta, N. T., & Castell, S. W. (2002). Adsorption, sequestration and bioaccessibility of As(V) in soils. Environmental Science & Technology, 36, 4562–4569. doi:10.1021/es011507s. Yoshida, K., Inoue, Y., Kuroda, K., Chen, H., Wanibuchi, H., Fukushima, S., et al. (1998). Urinary excretion of arsenic metabolites afterlong-term oral administration of various arsenic compounds to rats. Journal of Toxicology and Environmental Health. Part A., 54, 179–192. doi:10.1080/ 009841098158890.