• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Environmental Risks at Finnish Shooting Ranges-A Case Study

Posted on: Tuesday, 30 October 2007, 06:00 CDT

By Sorvari, Jaana

ABSTRACT We studied three Finnish shooting ranges in order to define the extent of the risks associated with elevated environmental concentrations of metals and PAHs. A scoring system revealed that lead, arsenic, and antimony were the most critical contaminants. On Site 3, the concentration of lead in groundwater exceeded the drinking water standard indicating evident health risks. For the remaining two sites we calculated Acceptable Daily Doses (ADD) based on the Reasonable Maximum Exposure (RME) approach and compared them with safe exposure levels. We also used a pharmacokinetic model to determine blood lead levels (PbBs). Risks to biota were assessed using ecological benchmarks and exposure and accumulation models. Prediction of leaching was based on laboratory tests and a distribution model. The health risk assessment for lead resulted in the maximum hazard quotient (HQ) of 1.2 whereas the HQs of As and Sb remained less than 1. Some exposure scenarios produced PbB estimates exceeding 10 [mu]g dl^sup -1^ but based on the uncertainty analysis we expect the health risks to remain insignificant. However, leaching of contaminants presents a risk to groundwater quality. At site 1 the ecotoxicity-based HQs demonstrate high risks to soil biota, small mammals, terrestrial plants and aquatic organisms.

Key Words: shooting range, contamination, risk assessment, ecological risk, health risk.

INTRODUCTION

Our recent inventory revealed that the number of outdoor shooting ranges in Finland totals 2,000-2,500 (Sorvari et al. 2006). At these, the level and extent of environmental contamination depend on the type of the shooting activity. In rifle shooting, bullets accumulate in natural barriers. When the bullets penetrate through the sandy backstop, abrasion causes an immediate environmental release of lead and other metals (Hardison et al. 2004). In the long term, ammunition decomposes due to weathering. Contamination on rifle ranges is manageable because of the treatability of sandy soil and limited area of pollution. Conversely, in shotgun shooting (i.e., skeet, trap, and sporting clays shooting) ammunition is spread around the range and contamination can extend beyond 10 hectares. At shotgun ranges, shot weathering is the main mechanism of emissions to soil.

Lead is the main component of traditional ammunition. It makes up about 97% in shot and about 90% in rifle bullets, depending on the type (e.g., Tanskanen et al. 1991; ITRC 2003). Lead is soluble only within a limited pH range and it tends to form poorly soluble compounds and bind to soil particles (Adriano 1986). Consequently, groundwater contamination by lead originating from ammunition has been reported only in a few cases (Soeder and Miller 2003; UMK-AG 1998).

Lead is poisonous in all forms and its effects in humans are diverse and depend on the level and duration of exposure (WHO Working Group 1995). Health effects are usually observed with blood lead (PbB) levels exceeding 1.93 [mu]mol l^sup -1^ (40 [mu]g dl^sup - 1^). Young children are the most susceptible to adverse health effects, for example, neurological and behavioral disturbances, resulting from lead exposure (e.g., HSDB 2003). In recent studies, Canfield et al (2003) discovered that in young children even blood lead levels less than 10 [mu]g dl^sup -1^ could cause lowering of IQ whereas Selevan et al. (2003) showed delayed growth and pubertal development in girls at the PbB level of 3 [mu]g dl^sup -1^.

Elevated environmental concentrations of lead can also give rise to ecological impacts. The toxicity depends on the type of soil and the chemical speciation of Pb. Thus, lead is generally more bioavailable and toxic in acidic than in alkaline or saline conditions. Furthermore, organic lead compounds are more toxic than inorganic compounds. In terrestrial plants, uptake of Pb can result in diminished growth and productivity, and decreased photosynthesis and water absorption (e.g., Bazzaz et al. 1974; Koeppe 1981). The lowest effective concentrations (LOECs) associated with phytotoxicity vary between 50 and 5,000 mg Pb kg^sup -1^ in soil depending on the plant species (Versluijs and Otte 2001). Adverse effects on terrestrial fauna include decreased and disturbed reproduction and growth, and increased mortality, among others (e.g., Statens Naturvardsvark 1984; Spurgeon et al. 1994; WHO 1989). The benchmarks (LOECs, NOECs, EC^sub 50^) based on the toxicity to earthworms and soil microbes range from about 400 to 5,000 mg Pb kg^sup -1^ in soil (Spurgeon et aL 1994; Efroymson et al. 1997a). The NOAELs of small mammals are typically two to three orders of magnitude lower compared with those of soil fauna. We can also expect lead accumulation in predators because small mammals serve as important prey for higher trophic levels. In an aquatic environment some organisms, for example, phytoplankton, filter-mollusks, bryophytes, and in lesser degree also fish and microcrustaceans, bioaccumulate lead (Thornton et al. 2001). However, biomagnification at higher trophic levels has not been reported.

Besides lead, traditional ammunition typically contain small amounts of antimony, arsenic, cadmium, copper, nickel, and zinc. Arsenic and antimony differ from the other metals in that in alkaline conditions they can form oxyanions, which are very mobile. Due to the higher mobility and higher human toxicity compared with lead, potential migration of As and Sb to groundwater used for domestic water intake can create a significant risk to human health. Equivalent to other metals in general, the toxicity and the toxic action of As and Sb depend on their physical state and speciation, that is the chemical form, valence and oxidation stage. Generally, the most soluble chemical species are the most toxic. Additionally, in acidic conditions strong reducing agents, such as metallic zinc, can reduce arsenic and antimony to highly toxic gaseous arsine (AsH^sub 4^) and stibine (SbH^sub 4^). Volatilization can also result from the microbial formation of arsene and its methylated forms (e.g., Bentley and Chasteen 2002). Some plants, algae and animals are also capable of biomethylation. Toxic manifestations of exposure to arsenic can appear in various organs and arsenic is also carcinogenic and genotoxic (ATSDR 2000). The symptoms of exposure to Sb resemble those of arsenic (Meditext(R) 2000). Moreover, delayed growth has been identified in children exposed to antimony (ATSDR 1992). The International Agency for Research on Cancer (IARC) has classified Sb^sub 2^O^sub 3^ as a potential human carcinogen. On the basis of ecological benchmarks (Efroymson et al. 1997a,b), arsenic, compared to lead, is more toxic to earthworms and soil microbes whereas Sb is more toxic to terrestrial plants and mammals.

In addition to metals, polyaromatic hydrocarbons (PAHs) used in clay disk targets are significant contaminants at shotgun ranges. The PAH concentration in clay targets varies considerably, that is, from 30 mg kg^sup -1^ to 40,000 mg kg^sup -1^ (expressed as a sum of 16 PAHs), depending on the manufacturer (UMK-AG 1998; Baer et al 1995). The compounds include probable human carcinogens, for example, chrysene, fluoranthene, and benzo(a)pyrene, among others.

Since the possible environmental risks associated with contamination at shooting ranges have been recognized, studies on shooting ranges have been performed in various countries during the past few years. The results of these studies show lead accumulation, adverse effects, and increased mortality on local terrestrial and aquatic biota representing different trophic levels (e.g., Lewis et al. 2001; Kendall et al. 1996; Stansley and Roscoe 1996; Roscoe et al. 1989; Ma 1989; Manninen and Tanskanen 1993; Scheuhammer and Dickson 1996). The major impact on wildlife is expected to resultfrom the unintended ingestion of lead shot (Thorn ton et al. 2001). Previous studies have also indicated serious health risks to children originating from ingestion of contaminated soil or bullet (Wixson and Davies 1994). Health risks can also arise from the contamination of food sources due to hunting with lead shot (Johansen et al. 2004), and from the exposure of cattle grazing on former shooting ranges (Braun et al. 1997).

In Finland, contamination at shooting ranges and possible risks related to it have not been studied systematically. Furthermore, previous site-specific risk assessments and risk management plans have mainly concentrated on lead only. This article presents results of a research project that comprised site studies and risk assessments on three Finnish shooting ranges. In this study, all key contaminants were considered. The study also included the planning of risk management actions and identifying data gaps and future study needs.

CHARACTERISTICS OF THE STUDY SITES

General Site Properties

The three shooting ranges we studied differed from each other in age and environmental conditions (Table 1). Both rifle and shotgun shooting had been practiced on all sites and all ranges were still in active use. All sites were situated on groundwater catchment areas belonging to the highest quality class, that is, class I^sup 1^ according to the Finnish classification. Groundwater was also used for domestic water intake at all sites. No endangered or protected species were present at any of the sites. However, Site 1 was situated adjacent to a protected natural area. None of the sites was particularly valuable either as a recreational area or as a hunting or fishing area. The most common soil type at all sites was coarse till or sand extending from the subsurface down to the groundwater level and below. The topsoil consisted mainly of a podzolic layer of varying depth with an ignition loss typically around 80 to 90% (corresponding to organic carbon content of about 50%).

Level of Contamination

Study methods

To find out the level and distribution of contamination, the soils at the study sites were studied onsite by x-ray fluorescence equipment (x-met) using a 40 m x 40 m sampling grid. At each site, the size of the impact area was determined on the basis of the estimates of shot flight path and the dimensions of the backstop (rifle ranges) (NSSF 1997). The estimates of the distribution of shots were based on the size of ammunition used as well as the location and the number of firing positions.

Composite soil samples were taken at different soil depths, and their total number varied between 113 (Site 2) and 223 (representing 137 sampling points, Site 1). Of these, selected samples representing the hot spot areas, that is, sampling points where a high concentration of Pb was identified by x-met, were taken for laboratory analyses. Here, the term "high concentration" refers mainly to the concentration above the limit value2 for soil. Groundwater samples were collected in summer and winter during three years at all study sites. Additionally, the sediment of the pond and surface water drains located at Site 1 were studied. At Site 1 and 2 samples of mushrooms and berries were collected in the hot spot areas (high concentration of Pb) and in a reference area. The total number of laboratory samples taken in different environmental media is presented in Appendix A, Table A1 and A2.

The selected soil samples, the sediment samples, and the samples of mushrooms and berries were taken to a laboratory where they were pretreated with concentrated nitric acid or aqua regia. The concentration of metals (Al, As, Cr, Cu, Fe, Ni, Pb, Sb, Zn) in the eluates and in water samples was analyzed by ICP (Inductively Coupled Plasma) equipment. In the case of surface water samples, both filtrated and unstrained samples were analyzed. Soil properties, such as pH, organic matter, particle size, and dry matter were determined from a few samples representing different soil types.

In addition to metals, PAHs were determined from one soil sample on Site 1 and from two soil samples on Site 3. These samples were taken within the outfall area of clay targets. The samples were pretreated according to the ISO 11464 standard (ISO 2006) and PAHs analyzed using the ISO 13877 method (ISO 1998).

Results

Pollution levels of the sites varied and there was no clear correlation between the temporal scope of shooting activities and the metal concentrations in the environment. On all sites, concentrations of lead and antimony exceeded the Finnish soil limit values and natural background concentrations at several sampling locations (Appendix A, Table A1). Furthermore, elevated concentrations of arsenic were discovered in a few samples.

According to the chemical studies, contamination at the shotgun ranges was mainly restricted to the topsoil extending from the surface down to 20 cm. However, elevated concentrations of metals were found at the depths of 40 and 45 cm at a few sampling points on Site 1 and Site 3. We also observed elevated concentrations of lead in the filtrated groundwater samples. However, only at Site 3 did the lead concentration exceed the quality standard for domestic water (10 [mu]g l^sup -1^). On Site 2, the contamination in soil was mainly limited to the depth of 5 cm from the ground level.

Concentrations of lead in the samples taken from lingonberries and mushrooms exceeded in some cases the Finnish guideline value of 0.1 mg kg^sup -1^-fw issued for foodstuff (Appendix A, Table A2). The sediment of the pond at Site 1 was heavily contaminated by lead. In addition, elevated concentrations of As and Sb were detected. The highest concentrations, however, were limited to the upper sediment layer (0-20 cm). High concentrations of particle-bound and dissolved lead were also detected in the water.

No PAHs were detected in the soil samples on Site 3 whereas on Site 1, the concentration of some PAHs, that is, chrysene, fluoranthene, benzo(6)anthrasene, and benzo(a)pyrene, exceeded the limit values for soil. Other PAHs, that is, dibenzo(o, h)anthrasene, benzo(k)fluoranthene, fluoren, phenantrene, anthrasene, and pyrene, existed in concentrations exceeding the lower guideline values but being less than the limit values. The high concentrations were presumably caused by some particles of clay pigeons, which were present in the samples. Similar problems have been reported in some previous studies abroad (Belway 2001). Due to the analytical difficulties and the limited number of analyses, it was not possible to draw any definite conclusions concerning the concentration of PAHs in the environment.

We did not find any distinct correlation between the results of x- met analyses and laboratory analyses. This is understandable because x-met results are strongly influenced by soil properties, such as particle size, moisture, heterogeneity, and the presence of other elements. However, an ordinal correlation was discovered between these two study methods. We also studied the correlation between the concentration of lead and the concentration of other metals. This examination was performed separately for different soil types. The ordinal correlation was rather strong between the concentrations of copper and lead, and between antimony and lead in sandy soil (r = 0.72, r = 0.77), but slightly weaker between zinc and lead (r = 0.61). Conversely, arsenic and nickel concentrations did not correlate with lead concentrations. In peat, only nickel showed a weak positive ordinal correlation with lead.

RISK ASSESSMENT

Scope of the Assessment

The aim of the site-specific risk assessments was to identify and assess current and future risks to humans and biota caused by the elevated concentrations of the key metals at the study sites. However, there was no need to run a separate quantitative health risk assessment (HRA) for Site 3 because the concentration of Pb in groundwater exceeded the drinking water standard. Hence, health risks at this site were already evident. Nevertheless, the ecological risks were assessed. For Site 1 and Site 2, the risks to groundwater quality were determined quantitatively. Due to the inaccuracy and insensitivity of the x-met method, the risk assessments were based solely on the results from the laboratory analyses.

According to some risk assessment studies, PAHs in addition to metals, could cause health risks at shooting ranges (Urzelai et al. 2003). However, we did not consider PAHs as critical contaminants in our study because the load from clay targets was limited to a small area, that is, about 0.3 hectares, corresponding to the outfall zone of clay targets. We also expected the PAHs present in clay pigeons to be practically immobile because they are mostly non-volatile and their solubility is very low. Some previous studies have actually shown that the PAHs in clay targets are strongly bound to the clay material (Crossman et al. 1989; Ikavalko 1997, unpublished). In the study by Crossman et al. (1989), only 1-18% of the PAHs were dissolved in leaching experiments performed on crushed disks. Additionally, Enell et al. (2004) reported a 0.3% removal of PAHs from soil material in leaching tests simulating natural conditions. Furthermore, in the study by Baer et al. (1995), the eluates from the extraction of clay targets were non-toxic to aquatic biota both in acute and chronic exposure. These studies strongly support our assumption of low mobility and low bioaccessibility of PAHs at our study sites. The mobility and bioaccessibility of PAHs was in fact also overestimated because acetone was used as a solvent in the pretreatment of samples. Dissolution in water is expected to be much lower.

Possible Contaminant Pathways

Some studies have shown that lead retention in Finnish forest soils varies from 26% to 54% of the total input to soil (Ulkomaanaho et al. 2001). Additionally, because organic matter binds metals effectively, we would expect the mobility of lead and most of the other metals at the study sites to be low and in turn their leaching to deeper soil layers and to groundwater to be minimal. However, due to the low pH in the upper soil layers, leaching downward is a potential pathway. Regardless of the fact that the environmental conditions, such as contamination level, pH and depth of groundwater table, were far more favorable for the leaching of lead on Site 1 compared with other study sites, we verified groundwater contamination by lead only on Site 3. Therefore, it seems highly probable that also at Site 1 contamination would in the future extend to the underlying aquifer.

In addition to leaching, run off proved to be a relevant pathway at Site 1. In fact, studies on the drains at this site showed high concentrations of lead (Appendix A, Table A2). Thus, some migration offsite may take place. However, the considerably lower Pb concentrations in the collective drain compared with the marsh area show an efficient dilution along the increasing distance from the most contaminated area.

In air, the migration of practically non-volatile substances such as lead, copper, nickel, and zinc, takes place mainly in paniculate form. The high concentrations of lead found in the samples of moss and lichen suggest that the soil-bound lead is in fact spread by the air. On the other hand, because at all the shotgun ranges and at the rifle range on Site 2, the soil is densely covered with vegetation about 180 days a year and with snow for about 160 days a year, we expected soil erosion and thus contaminant transport via suspended particles in the air to be minimal. In fact, the elevated concentrations of lead in moss and lichen can originate from metal- containing fumes and particles generated in the actual shooting event (Tukker et al. 2001; NSSF 1997). Unlike lead and the other metals, arsenic and antimony can migrate in gaseous form due to their tendency to transform into volatile compounds in some environmental conditions (see Introduction). Although such conditions probably exist particularly at Site 1 such volatilization was considered as insignificant in respect of exposure due to the facts that the present and future land use would not cover indoor activities, and that the elevated As and Sb concentrations were mainly concentrated on a few sampling points. In practice, these reactions require existence of reducing conditions and the presence of a reducing agent, for example, metallic zinc. These prerequisites were fulfilled on Site 1. However, especially because the present and future land use would not cover indoor activities, and because the elevated As and Sb concentrations were mainly concentrated on a few sampling points, we considered volatilization of As and Sb as insignificant in respect of exposure.

Materials and Methods

Identification of critical substances

We used a simple scoring system presented by the United States Environmental Protection Agency to identify the most critical contaminants. This method is based on the calculation of risk indexes by multiplying the toxicity of each contaminant by its environmental concentration (USEPA 1989). In the case of identification of health risks, the inverse of the lowest and meaningful reference doses (RfDs) served as a measure of toxicity. Risk indexes were calculated for the following exposure routes: drinking water, soil, and edible plants. Outdoor air was excluded because there were no data available on the concentrations at the study sites. To identify contaminants relevant to local biota, we used several sets of ecotoxicity-based benchmarks (BM) derived for different organisms (Table 2).

On the basis of the scoring method, lead, arsenic, and antimony proved to be the most significant metals associated with both health risks and ecological risks on all exposure routes. Of these, lead and arsenic had the highest risk scores. The risk indexes of copper were also high in some cases (some ecological receptors). The environmental concentrations of nickel were generally low, that is, less than the soil guideline value and ecological benchmark values. Consequently, its risk indexes were from one to three orders of magnitude lower compared with the highest risk indexes. Therefore, nickel was excluded from further risk assessment.

Health risk assessment (HRA)

No changes in land use were expected at any of the study sites. Therefore, in the HRA we considered the following land-use scenarios: (1) use as a shooting range, (2) use as a domestic water resource (groundwater abstraction), and (3) use as a recreational area. Recreational use included outdoor activities, such as sports and collection of edible plants. Taking into account all these scenarios, human exposure to the contaminants could occur through breathing of air dust, dermal intake, soil ingestion, consumption of local foodstuff (mushrooms, berries), and through the use of tap water.

We employed a tiered approach in the HRA. In Tier 0, a simple qualitative classification system was used to identify relevant pathways and exposure routes. For these relevant pathways and routes we performed a Tier I assessment and calculated average daily doses (ADDs) (Eq. 1).

Contaminant-specific hazard quotients (HQs) were determined as a ratio of ADD to the respective acceptable daily intake (RfD /TDI, ADI). In the case of lead, we also used a pharmacokinetic model to determine the blood lead level (PbB) (Eqs. 2 and 3). The methodology based on the use of reference doses or tolerable/acceptable daily intake values (TDIs/ADIs) is used commonly in lead risk assessment in Europe, whereas in the United States and Canada, site-specific risk assessments associated with lead exposure are based on the evaluation of PbB. The United States Environmental Protection Agency (USEPA) has recommended using the adult lead uptake model in the case of land use scenarios other than residential areas (IJSEPA 1996a). The model specifically estimates accumulation of lead in the blood of women of the childbearing age and their unborn children. For unborn children (fetuses) we used only the PbB model to assess the risks.

We did not consider young children between the ages of 1 and 6 separately although they are expected to show higher soil ingestion rates than adults and children of other age groups. However, based on the land-use scenarios we did not expect small children to visit on the study sites. Furthermore, risks to remediation workers were estimated only on the basis of the concentrations of the key metals in air, thus no quantitative exposure estimates were derived.

The quantitative Tier I risk assessment was based on a reasonable maximum exposure (RME) approach. Therefore, for parameters describing human exposure we chose the 90th or the 95th percentile values when available as suggested by the USEPA (1989, 1997). From the concentrations of the key metals detected in the soil and sediment samples we calculated the 95% upper confidence limits (UCLs) of the arithmetic means in accordance with the recommendations issued by the USEPA (1989, 1992). All parameters included in the equations and their values are shown in Table 3.

Ecological risk assessment (ERA)

The exposure routes of terrestrial fauna are practically the same as in the case of humans. The drinking water route, however, is mainly associated with surface water. In the case of terrestrial plants, wet and dry depositions are important pathways in the uptake of contaminants through plant parts above ground. Direct intake by roots from soil is expected to be low for substances with low solubility (e.g., lead, copper, nickel, zinc). In fact, a previous study on lead accumulation in plants on a Finnish shooting range showed concentrations in plants being about 1,000 times lower compared with the concentration in humus (Manninen and Tanskanen 1993). Furthermore, uptake of lead has normally been confined to root tissues (McGeer et al. 2004). Consequently, significant accumulation would be expected only in tuberous plants, for instance potatoes and carrots. However, in contrast to these findings, a recent study by Finster et al. (2004) showed that some leafy vegetables and herbs are also able to take up lead from the soil to their stem and leaves. Consequently, it is necessary to consider the accumulation of the key metals in leafy plants as a potential risk factor.

The available data on concentrations of contaminants in local biota at our study sites comprised results from a very few studies of mushrooms, berries, moss, and lichen. Furthermore, no surveys on the type, diversity, or size of populations were compiled. Therefore, the ecological receptors were selected on the basis of the type of landscape and knowledge of their typical fauna; availability of toxicity data, for example, LOAELs, NOAELs, EC-/LC- values; sensitivity to the key contaminants; ecological importance of the species at ecosystem level; and suitability of scale. Hence, we excluded terrestrial birds and large mammals, for example, hares, because the total habitat of these animals is large compared with the surface area of the study sites. On the contrary, earthworms and shrews meet our selection criteria. To estimate bioaccumulation in these key species, we used validated models, which are based either on uptake factors (UF) or simple linear regression and exposure models (Table 4). However, no quantitative model was available for the assessment of bioaccumulation of antimony in earthworms. The concentrations used in the calculations were selected according to the organism in question, that is considering its habitat (Table 4). We also followed the principles of the conservative, screening level risk assessment. Here, the recommendations and information presented by Sample et al. (1998a,b) were adopted.

Earthworms were actually not the most meaningful receptors to consider because they do not dwell in the typical Finnish forest soil. Thus, rather than serving as receptors per se, we assumed earthworms represented the extent of the exposure of other common soft-bodied organisms typical in Finnish moor areas, such as Enchytraeidae, and also accumulation in the food web. Here, shrews were used as an indicator species assuming that they feed solely from earthworms. In addition to the assessment of accumulation of the critical metals in shrews and earthworms, we calculated HQs for soil microbes, terrestrial plants, earthworms, and some aquatic organisms by dividing the environmental concentrations with the species-specific benchmarks (see Table 4).

Some migratory waterfowl were observed during site investigations at Site 1. Due to the low number of individuals and their short dwelling time, we did not consider risks to waterfowl in the ERA. However, it was recognized that the few individual birds are at risk because of unintentional shot ingestion.

Evaluation of contaminant transport

The evaluation of speciation, solubility, and migration of lead in soil was based on different leaching tests using different liquid to solid ratios (L/S). These included extraction with NH^sub 4^Ac- EDTA (pH = 4.5), the availability test (two stages, pH = 7 and pH = 4, L/S = 50 + 50), and the Dutch percolation test NEN7343 (cumulative L/S = 10). The extraction with NH^sub 4^Ac-EDTA indicates the relation between dissolved and metallic lead in soil whereas the availability test simulates the worst case situation, that is, the maximum teachability. The percolation test aims at describing the actual leaching as a function of time. The results of the latter test can be extrapolated to give an estimate of the long- term leaching behavior (Eqs. 4, 5, 6). The percolation tests and availability tests were run site-specifically with two samples representing the two dominating soil types, that is, sand and peat. To compare the lead concentration measured in the eluate (E^sub max^) with the target pore water concentration (C^sub r^), the results from the leaching tests were divided by the L/S-ratio used in the leaching test.

Because the concentration of critical metals other than lead were not determined from the eluates, the soil pore water concentrations were estimated using the distribution coefficients presented in the literature. Following the conservative approach, in Tier I we ignored possible dilution and assumed the groundwater concentration below the contaminated soil to equal the concentration in soil pore water (Eq. 7).

In the Tier II assessment, dilution coefficients were taken into account (Eqs. 8 and 9). The national quality standards for domestic water served as groundwater target concentrations.

The parameters used in Eqs. 4-9 and their values are described in Table 5.

We utilized the estimates of the total amount of lead on site, information on the active age of the range, and the results from soil solubility tests using NH^sub 4^Ac-EDTA as a solvent to estimate the rate of corrosion of lead shot. Here, we assumed that the shot load and the shot decay rate were both constant during the lifetime of the shooting range. Thus as a simplification, we ignored possible formation of secondary minerals, which could produce an insoluble layer on the shot surface and thereby prevent corrosive substances from coming into contact with the metallic lead (Lin et al. 1995; Lin 1996; J0rgensen and Willems 1987).

Transport of contaminants along the drains (Site 1) was assessed only qualitatively because there were no data (e.g., on the sedimentation rate, resuspension, flow rate) available for quantitative assessment and because site studies showed efficient dilution alongside increasing distance from the hot spot areas.

RESULTS

Health Risks

Using the lowest acceptable daily dose of 1.0 [mu]g Pb kg^sup - 1^d^sup -1^ (UMS 1997) documented in the literature studied, the tier IHRA resulted in HQ^sub Pb^ values of 1.2 (Site 1) and 0.9 (Site 2). Use of a more recent reference dose of 3.6 [mu]g Pb kg^sup -1^ d^sup -1^ (Baars et al. 2001) resulted in HQ^sub Pb^ values of 0.3 and 0.2, indicating insignificant risks at both study sites. The higher reference value has been generally adopted in Finland and it has also been used as a basis in the determination of the new soil guideline values. These risk estimates refer to exposure through all possible routes and they also include the average Finnish background exposure resulting from food consumption. Here we used the upper estimate 0.020 mg d^sup -1^ presented by Alfthan (1994).

The results of the HRA also show that the division between different exposure routes is different at Site 1 from that at Site 2. At Site 1, soil ingestion is the main exposure route covering about 40% of the total exposure whereas at Site 2, food ingestion (local mushrooms and berries) is clearly the main exposure route corresponding to 55% of the total exposure. These differences are due to the differing concentrations of lead in berries, mushrooms, and soil as well as to unequal exposure times. At both sites other exposure routes cover less than 10% whereas the proportion of the background exposure is about 30% of the total exposure.

The results of the calculations based on the PbB model demonstrate that the blood lead level of a fetus might reach a level of 10 to 20 [mu]g dl^sup -1^ if the mother had been exposed to lead at Site 1 or Site 2 for several years and through all routes of exposure. In children, the USEPA and the U.S. Food and Drug Administration (FDA) consider 10 [mu]g dl^sup -1^ to be a safe level. However, some studies have shown health effects at considerably lower concentrations (see the Introduction for more information). The Finnish reference value for a non-exposed pregnant female is 0.3 [mu]mol l^sup -1^ (= 6.2 [mu]g dl^sup -1^) (Frilander and Taskinen 1999). The most conservative exposure scenario resulted in a maximum PbB level of 9 [mu]g dl^sup -1^ and hence, the national reference value is exceeded about 1.5-fold.

Food intake seems to be an important route of lead exposure at both sites. Therefore, elimination of this pathway by issuing restrictions for recreational activities would diminish exposure effectively. Impeding the intake of local foodstuff, that is, berries and mushrooms, would clearly decrease the ADDs of lead and PbB levels less than the maximum permissible levels.

The HQs of arsenic and antimony were well less than 1 at both sites, that is, 0.03 (As) and 0.1 (Sb) in the case of Site 1 and 0.02 (As) and 0.02 (Sb) in the case of Site 2, respectively. These HQ estimates are based on the reference doses given in the database of the Integrated Risk Information System (IRIS, http://www.epa.gov/ iris), that is, RfD^sub As^= 3*10^sup -4^ mg kg^sup -1^ d^sup -1^ and RfD^sub Sb^ = 4*10^sup -4^ mg kg^sup -1^d^sup -1^. The dose estimates of As and Sb refer to all possible site-specific exposure routes but they do not include exposure from diffuse sources (the background), such as food, since there was no such information available.

According to the methodology issued by the U.S. Agency for Toxic Substances and Disease Registry (ATSDR 2001), the joint toxic actions of chemicals should be considered if the individual HQ value of at least two contaminants exceeds the value of 0.1 at the first assessment level. Because this was not the case, we expected the combined effects of the studied metals to be insignificant.

The estimated concentrations of Pb, As, and Sb in respiratory air were clearly less than the maximum acceptable concentrations at work place issued by the Ministry of Social Affairs and Health (2002). Therefore, occupational exposure during remediation is not expected to result in health effects.

Ecological Risks

The results of the ERA show that some adverse ecological effects are expected at all study sites. However, these results are to be considered as provisional only because they include many uncertainties (see section Uncertainty Analysis below).

The HQs we calculated indicate ecological risks at Site 1 in particular. Lead is clearly the most critical contaminant. The site- specific HQ^sub Pb^ values vary between 0.4 and 7 (Site 3), between 0.8 and 15 (Site 2) and between 15 and 260 (Site 1) depending on the receptor (microbes, earthworms, plants). At Site 1, the HQ associated with phytotoxicity of antimony is also high, that is, 240. The accuracy of these results is reduced by the fact that we used the maximum concentrations detected in the calculation of the HQs.

Although lead proved to be the key contaminant causing risks to plants, earthworms, and microbes, in the case of shrews as receptors, the major toxic response at the shotgun range at Site 1 is expected to be caused by antimony as shown by the HQ^sub As^ = 2 and HQ^sub Sb^= 10. At Site 2, the concentrations of As and Sb were low resulting in low HQs ([much less than]1). At the same time, the estimated HQs of lead are 0.2 (Site 2) and 3 (Site 1) in the case of shotgun ranges and 4 (Site 2) and 11 (Site 1) in the case of rifle ranges.

The benchmarks stated for lead in sediment vary between 15 and 396 mg kg^sup -1^ (Efroymson et al 1997a; Jones et al 1997; CCME 1999). At Site 1, the average concentration of lead in the sediment exceeded these values by almost 400-fold at the maximum. The highest concentrations of arsenic and antimony were also about 6-and 40- fold higher compared with the lowest sediment benchmarks. Additionally, the average concentration of dissolved lead exceeded the concentrations toxic to aquatic biota, except for aquatic plants. Hence, we can expect high ecological risks to aquatic biota.

Risk of Groundwater Pollution

In the case of Site 1, the availability test showed that the maximum solubility of soil-bound lead was 10% in the acidic, peat soil and 1.5% in the sandy soil. Assuming no dilution takes place, in groundwater leaching of Pb would lead to about 1,000-fold and 100- fold exceeding of the quality standard for domestic water (0.01 mg l^sup -1^) at the maximum and depending on the soil type. The solubilities in the availability tests were about 10-fold higher compared with the results from percolation tests. Assuming a log- linear leaching behavior, the extrapolation of the results from the percolation test (Eqs. 6, 7, and 8) produced time estimates of 70 to 100 years in the case of sandy soil, and of 6.5 to 20 years in the case of peat soil (Figure 1). These estimates refer to 4 different sub-samples of soil and they illustrate the time that it will take for all the soil-bound lead above the aquifer to reach the groundwater table. In the marsh area, where the pH is low and the groundwater table is near the ground level, leaching would be clearly the quickest. All these results represent high-end estimates because they are based on conservative values of the depth of groundwater table. In practice the distance to ground level varied considerably, that is, between 0.09 m and 7.5 m, in the measurements run in late spring.

It should be noted that the corrosion of lead shot was ignored in the estimates of leaching potential. Lead shots form a stock from which lead and other metals are dissolved continuously. In our study, we ended up at estimates of annual shot corrosion rates between 0.1 and 1.4% depending on the site and soil properties. These results show that even if shotgun activities ceased at shooting ranges, in the marsh area it would take about 70 years for all the metallic lead to dissolve whereas in the sandy soil the total dissolution could take 1,000 years. The estimated concentrations of As and Sb in the leachate exceed the domestic water quality standards. In the case of arsenic, the highest estimate of leachate concentration at Site 1 is about two orders of magnitude higher compared with the quality standard (0.01 mg l^sup - 1^) depending on the hydraulic conductivity of soil. In the case of antimony, the highest concentration estimate exceeds the quality standard (0.003 mg l^sup -1^) with a factor of about 1,000. At Site 2, the risks to groundwater quality are considerably lower due to lower concentrations in soil and conditions less favorable to leaching (e.g., higher pH). It has to be noted that our calculations were based on rather conservative parameter values, resulting in very low dilution coefficients (DF varying between 1 and 4 on Site 1) and pore water concentrations. These and other factors causing uncertainty and possible overestimation in the assessment of risks to groundwater quality are described in detail in the section that follows.

Uncertainty Analysis

Overview of the factors causing uncertainty and variability

The quantitative risk assessments included many uncertainties arising from the assumptions and assessment methods and natural variability. Because the conservative HRA resulted in low HQs, the uncertainty was assessed only qualitatively. Furthermore, because we had no statistical data concerning the parameters used in the ERA and in the assessment of risks to groundwater quality, we could not use quantitative methods in the characterization of ecological risks and risk of groundwater pollution. Results of the qualitative uncertainty analysis are presented in Table 6.

The use of the PbB model lead to rather high risk estimates, that is, high PbB levels of unborn children in particular. These results cannot be compared straightforwardly with the HQ^sub Pb^ values because children were not considered as receptors in the calculation of ADDs. Moreover, in the case of Site 1, to calculate the ADDs a lower exposure time (30 d a^sup -1^) compared with the calculation of PbBs (90 d a^sup -1^) was used in the determination of exposure through soil ingestion and inhalation. The use of an equivalent frequency of exposure would lead to HQ^sub Pb^ = 0.6 when using the higher reference dose and HQ^sub Pb^ = 2.3 when using the lowest reference dose. Taking into account the environmental conditions such as the length of snow cover period, the location of Site 1, and the land use, it is quite improbable that the actual time any individual would be exposed to the contaminants of the site would be more than 30 days in a year. Due to the same factors the estimate of EF = 90 da^sup -1^ can be considered as conservative in the case of Site 2. The high estimates of exposure time were used because the PbB model is actually suitable only for cases in which the ED >/=90 days per year. Furthermore, the probability of the scenario that the possible receptor would be maximally exposed through all possible routes (food intake, soil ingestion, consumption of drinking water, inhalation of soil particles, skin contact) can be considered as low.

At both study sites, the value we used for the human exposure to lead originating from other sources (i.e., the background) makes up about 25-35% of the estimated daily exposure and the PbB^sub adult^ values. Hence, the background values have a considerable impact on our risk estimates. Here, we used the Finnish estimates for the maximum daily dose associated with food intake and for the average PbB level. Such estimates were presented in the mid-1990s (Alfthan 1994). Some studies have reported about 10-fold higher daily intake values, that is, 231 [mu]g Pb d^sup -1^ for Finnish adult males and 178 [mu]g Pb d^sup -1^ for adult females (HSDB 2003). However, these higher PbB estimates date back to the time when leaded gasoline was still used in car engines. According to Ponka et al (1991), the average PbB level of children in daycare decreased about 30%, that is, to avalue of 3.0 [mu]g dl^sup -1^, between 1983 and 1988. This decline is due to the lowering of the environmental release of Pb. On the basis of these studies and due to the continuing declining tendency of lead emissions, we can expect the average daily lead exposure of Finns to be even lower at present. Consequendy, the PbB levels would also be lower. Hence, even the background values presented by Alfthan are probably conservative.

Use of the UCLs corresponding to the 95% confidence limit of the mean resulted in the high end estimates of risks. Because of the lower sensitivity and poorer accuracy of the x-met method, risk assessments were based solely on the results of laboratory analyses. Because the laboratory analyses were focused mainly on the hot spots identified on the basis of x-met analyses, the actual average concentrations at each site were overestimated. In the case of other metals than lead, the significant variation of the concentrations together with the limited number of samples resulted in high standard deviations of the means and in turn, high UCL values. Here, additional site studies, that is, laboratory analyses of samples representing the whole impact area, would have provided data for more accurate risk estimates.

In ERA we ignored the possible direct shot ingestion by shrews. Various studies dealing with shot ingestion by waterfowl have been reported previously. However, we were not able to find information on studies focused on small mammals.

In the assessment of the leaching of contaminants to groundwater, the parameters describing soil properties are the most critical. In our study the most uncertain factor was the hydraulic conductivity (k-value) because we had to assess its value solely on the basis of soil type. In practice, the variation in this parameter can be several orders of magnitude even if the soil texture is similar. Hence, it is important to determine the proper site-specific k- value empirically. The distribution coefficient (K^sub d^), which was used for the calculation of the concentration of arsenic and antimony (and lead in the case of Site 2) in pore water, is also an uncertain parameter and should be assessed site-specifically.

Bioavailability aspects

Biovailability is an important uncertainty factor both in HRA and ERA because low bioavailability restricts the absorption of contaminants in the receptor's body. In the calculation of human PbB, we used the default absorption factors (AF) issued by the USEPA (1996a), that is, 0.12 for lead in inhaled soil particles and 0.2 for lead in edible plants and drinking water. Conversely, following the RME mediodology we assumed 100% absorption on all exposure routes in the calculation of HQs. Furthermore, the excretion and metabolism of the contaminants in the body were not considered due to the lack of quantitative data, and also because the level of the quantitative HRA (screening). These assumptions will obviously lead to conservative health risk estimates.

In the calculation of exposure, particle size was ignored. Particle size is an important factor especially in inhalation because only particles smaller than 10 [mu]m can penetrate to pulmonary tissues. Moreover, pulmonary absorption decreases as particle size and density increase and solubility and surface area decrease. Also, the deposition of particles in air depends on their size; that is, the deposition decreases with the increasing particle size (NEPI 2000). Because small particles (<10 [mu]m) constituted only 5 to 15% of the particles in the surface soil on our study sites, the respiratory exposure was overestimated. On the other hand, exposure to inhaled metals has seldom been considered to be a major exposure route. The results of our study also show that the exposure through inhalation of particles is probably less than 10% of the total exposure. Therefore, ignoring of the limited bioaccessibility associated with the particle size distribution does not cause significant error in our final results.

In their study on shooting ranges, Turpeinen et al (2000) discovered that there is no straightforward correlation between the acid-soluble or water-soluble lead concentration and magnitude of ecological risks. In practice, only ionic lead (Pb^sup 2+^) and small-sized organic lead complexes are bioavailable. Furthermore, Turpeinen et al. report that in their study only 4-6% of the lead in peat soil was bioavailable to soil microbes. In mineral soil the part of the bioavailable lead was 13-43%. In our study, we analyzed the water-soluble lead using the availability test. In the soil samples collected from the most contaminated site (Site 1) only about 10% of lead was water-soluble in the peat and about 1 % in mineral soil, respectively. According to these data and information on the concentrations at study sites, the actual lead concentration bioavailable to soil microbes at the most polluted site, namely Site 1, would therefore be considerably lower, that is, less than 100 mg kg^sup -1^. Hence, we expect also the concentration in other soil organisms and shrews to be lower than estimated.

CONCLUSIONS AND DISCUSSION

Risks and Risk Management Actions at Studied Shooting Ranges

Our results from the quantitative HRAs show that the human health risks are low at Site 1 and Site 2, whereas ecological risks are at least moderate. However, the HQs describing ecological risks include significant uncertainties and they can generally be considered only as preliminary indicators of ecological risks. In fact, the uncertainty analysis revealed that the results of both HRA and ERA overestimate actual risks. Unlike at Site 1 and Site 2, health risks are apparent at Site 3 due to the verified contamination of groundwater used as a drinking water source. Our studies revealed that soil contamination at Site 1 presents a particular risk to groundwater quality due to continuous leaching of contaminants in the acidic marsh area. This observation is fully consistent with the general understanding of the behavior of lead and the findings of some previous studies concerning lead mobility at shooting ranges (Turpeinen et al 2000; Knechtenhofer et ai 2003; ITRC 2003; USEPA 2001). Besides acidity, an additional factor forcing contaminants to migrate to groundwater is the occasionally high groundwater table. At Site 1, we observed that the water table was almost at the level of the soil surface during springtime when snow is melting. Flushing from the highly contaminated topsoil is therefore possible.

Although the conservative HRA proved that the present level of contamination at Site 1 and Site 2 does not cause significant health risks, some simple and feasible risk management measures were proposed in order to minimize human exposure. For example, warning signs were mounted to inform people of contamination of local berries and mushrooms. Furthermore, no more groundwater was abstracted at Site 1 and an alternative water resource was introduced in order to eliminate possible future health risk caused by consumption of contaminated drinking water. To restrict future contamination, shotgun-shooting activities were also concluded. Because no actual remedial measures were initiated at this site, the site should be monitored in order to ascertain that the pollution of groundwater is prevented in the future. Actions to restrict ecological risks at Site 1 and Site 2 are still pending because the need for risk management actions and the feasibility of different management options need to be studied in detail. At Site 3, remediation was launched due to verified groundwater pollution.

Clearly, the most important long-term environmental risk at shooting ranges is the continuing corrosion of ammunition. According to our studies more than 50% of lead can still be in solid phase (in shot) after about 30 years of shooting activities. Therefore, lead pellets in soil form a stock from which metals continue to dissolve and pollute the environment. Estimation of the time scale of this decay is difficult due to the complex chemical behavior of metals in different environmental conditions. We ended up with estimates of 0.1% and 1.4% annual decomposition depending on the soil type. To compare these figures with some previous estimates, in a Danish study the estimated time of total decay was 100 to 300 years depending on the environmental conditions (Jorgensen and Willems 1987). In a study of Swedish soil, Zhixun et aL (1995) found that in 20-25 years 4.8% of metallic lead was turned into carbonate and sulfate compounds. In the same period, in peat the average fraction of the reacted lead was 15.6%. The sole previous Finnish study produced a decay estimate of 0.5% per year (Esko Rossi Oy 2001, unpublished). Thus, we can conclude that our results are in line with the results from other studies on shooting ranges. Accordingly, at our study sites it could take centuries or even a millennium until all contaminants have been released from the ammunition into the environment.

Uncertainties of Risk Assessment

Overall, the methods used in the risk assessments resulted in rather conservative, that is, high end, estimates of the total risks to human health, biota, and groundwater quality (see the section Uncertainty Analysis and Table 6 for a detailed description of the uncertainties). Additional studies would have provided data for a more realistic risk assessment, especially in the case of ecological risks and risks to groundwater quality. These study needs are described under the section Recommendations and Future Study Needs.

Methodologies to assess health risks at contaminated sites also vary in different countries. Despite the differences in the actual quantitative risk assessment procedure, both the European methodology based on the use of a reference dose (TDI) as well as the PbB method adopted particularly in the United States, are based on the same principle, that is, they use the acceptable PbB level of 10 [mu]g dl^sup -1^ of children as a basis. However, in our study case, the model for the assessment of PbB level proved to be slightly unsuitable because it is only applicable to situations where the annual time of exposure exceeds 90 days in all exposure routes, a situation that is improbable at our study sites. Moreover, the suitability of default parameters in the PbB model could not be verified because there were no data available on the values representative of Finnish conditions.

In addition to the variation in the risk assessment methods, the reference doses and concentrations used in the characterization of risks vary. Because this can cause significant variation in the results, attention should be paid on the selection of appropriate and up-to-date reference values.

RECOMMENDATIONS AND FUTURE STUDY NEEDS

Although the results presented refer to the three shooting ranges studied, they can be generalized in some extent to many Finnish shooting ranges. First of all, according to a national survey approximately one third of all Finnish shooting ranges are at least partly situated in aquifer areas, that is, sandy esker areas often used for the intake of domestic water (Sorvari et al 2006). The environmental conditions, for example, the prevailing soil types (moraine, sand, podzolic top layer) and marsh areas, characteristic to the study sites are also very typical in Finland. However, it must be stressed that the actual contaminant load is highly dependent on the age of the shooting range, the type of activities, and the number of shots fired. Therefore, these factors should always be determined site-specifically.

Due to limited biological studies at the study sites, the question of the scope and magnitude of ecological risks on Finnish shooting ranges remains. Because HQs are mainly based on a conservative starting point ignoring many relevant aspects, such as bioavailability, adaptation, recovery, and avoidance, among others, the presence and magnitude of ecological risks should be verified with bioassays and/or onsite monitoring of biota before the realization of extensive remediation actions. In fact, some Finnish studies on areas contaminated by metals show that due to these confounding factors, the actual risks at sites contaminated by metals are lower than expected (e.g., Pennanen et al. 1996).

Excluding the assessment of transport of contaminants to groundwater, our studies on migration of the key metals were limited to qualitative assessment based on site data. Possible migration at Site 1 in particular should be studied further in order to identify existing and future risks to the nearby protected marsh area. In the case of risks to groundwater quality, the prediction of the leaching of contaminants proved to be difficult solely on the basis of the knowledge on the chemical composition, generic hydrogeological conditions, and soil properties. The results from simple modeling, for example, based on equilibrium partition, in particular, merely serve as very preliminary information on possible risks. Despite favorable conditions for leaching of metals, that is, low pH and shallow groundwater table, groundwater may remain unpolluted whereas contamination in less favorable conditions may occur. Therefore, more detailed on site or laboratory studies on the mobility of contaminants, that is, leaching tests and lysimeters, and more reliable modeling of hydrogeological conditions should be performed in order to arrive at more realistic estimates of risks to groundwater quality. Above all, due to the difficulties to assess and manage the risks to groundwater quality, the shooting activities should be restricted or distribution of ammunition limited in the shooting ranges situated in the vicinity of important aquifers.

Our study proved that besides the prominent contaminant, namely lead, antimony, and arsenic can cause a significant risk to groundwater resources, terrestrial biota, and aquatic ecosystems. These contaminants, however, have usually been ignored in the previous case-specific risk assessments performed in Finland. Studies from some other countries also point to the importance of antimony and arsenic in the appearance of human health risks at shooting ranges (Kantonales Laboratorium Aargau 2002; Dynamac Corporation 1998). Hence, in the future these contaminants should be considered in the risk assessment and included in the planning of risk management actions at shooting ranges.

ACKNOWLEDGMENTS

This study was financed by the Finnish Ministry of the Environment, the support of which is gratefully acknowledged. The site studies and x-met analyses were performed by the project partners in the North Karelia Regional Environment Centre (Petri Naumanen andjari Tiainen) and West Finland Regional Environment Centre (Paivi Rajala). The contribution of these partners is highly appreciated. The laboratory analyses were run at the Geological Survey of Finland and Tampere University of Technology.

1 An aquifer belongs to the class I if its water is used by waterworks that provides or will within 20-30 years provide water for more than 10 households or the area is needed to provide domestic water in the case of crisis.

2 Finnish guideline values comprise two sets of substance- or compound-specific concentration limits. If all concentrations in soil are less than the lower guideline values, soil is considered to be clean. The higher values, that is, the limit values, indicate probable risks to humans. In most cases exceeding of any limit value necessitates some risk management actions (Jeltsch and Pyy 1994).

REFERENCES

Aalbers ThG, de Wilde PGM, Rood GA, et al 1996. Environmental Quality of Primary and Secondary Construction Materials in Relation to Re-use and Protection of Soil and Surface Water. RIVM 771402007. National Institute of Public Health and the Environment (RTVM), Bilthoven, The Netherlands Adriano DC (ed). 1986. Trace Elements in the Terrestrial Environment. Lewis Publishers, Boca Raton, FL, USA

Alfthan G. 1994. Lyijyaltistus vahentynyt selvasti (Considerable Decrease in Lead Exposure) . Kansanterveyslaitoksen tiedotuslehti 5. National Public Health Institute. Available at http://www.kd.fi/ portal/suomi/julkaisut/kansanterveyslehti/lehdet_1994/5.1994/. (in Finnish)

ATSDR (Agency for Toxic Substances and Disease Registry). 2001. Guidance Manual for the Assessment of Joint Toxic Action of Chemical Mixtures. Draft for public comment. US Department of Health and Human Services, Atlanta, GA, USA. Available at http:// www.atsdr.cdc.gov/interactionprofiles/ipga.html

ATSDR. 1992. Toxicological Profile for Antimony. Available at http://www.atsdr.cdc.gov/toxprofiles/

ATSDR. 2000. Toxicological Profile for Arsenic. Available at http://www.atsdr.cdc.gov/toxprofiles/

Baars AJ, Theelen RMC, Janssen PJCM, et ai 2001. Re-evaluation of Human-toxicological Maximum Permissible Risk Levels. RTVM 711701 025. National Institute of Public Health and the Environment (RTVM), Bilthoven, The Netherlands

Baer KN, Hutton D, Boeri RI, et al 1995. Toxicity evaluation of trap and skeet shooting targets to aquatic test species. Ecotoxicology 4:385-92

Bazzaz FA, Rolfe GL, and Windle P. 1974. Differing sensitivity of corn and soybean photosynthesis and transpiration to lead contamination. J Environ Qual 3:156-8

Belway S. 2001. Site assessment Union mine gun range Eldorado County, California. Kleinfelder, Inc. Sacramento, CA, USA

Bentley R and Chasteen TG. 2002. Microbial methylation of metalloids: Arsenic, Antimony, and Bismuth. Microbiol MoI Blol Rev 66(2).-250-71.

Braun U, Pusteria N, and Ossent P. 1997. Lead poisoning of calves pastured in the target area of a military shooting range. Schweitz Arch Tierheilkd 139:403-7

Canfield RL, Henderson CR, Cory-Slechta DA, et al 2003. Intellectual impairment in children with blood lead concentrations below 10 microg per deciliter. N Engl J Med 348:517-26

Carlon C, Dalla Valle M, and Marcomini, A. 2004. Regression models to predict water-soil heavy metals partition coefficients in risk assessment studies. Environ Pollut 127:109-15

CCME (Canadian Council of Ministers of the Environment). 1999. Summary of Existing Canadian Environmental Quality Guidelines. Available at http://www.ccme. ca/publications/ can_guidelines.html#110

Crossman G, Fahrehorst C, Klother G, et al 1989. Die Belastung von Boden auf Sportschiessplatzen durch Bleischrot und Wurftauben (Contamination of shooting range soils by lead pellets and clay pigeons). UBA-FB 89-100, Texte 35/89. Umweltbundesamt, Berlin, Germany (in German)

Dynamac Corporation. 1998. Final Engineering Evaluation/cost Analysis Sabino Canyon Shooting Range, Coronado National Forest, Arizona. Available at http://www.fs.fed.us/r3/coronado/eeca/

Efroymson RA, Suter GW II, Sample BE, et al 1997a. Preliminary Remediation Goals for Ecological Endpoints. ES/ER/TM-162/R2. US Department of Energy, Washington, DC, USA. Available at http:// www.esd.ornl.gov/programs/ecorisk/documents/tml62r2.pdf

Efroymson RA, Will ME, and Suter GWII. 1997b. Toxicological Benchmarks for Contaminants of Potential Concern for Effects on Soil and Litter. Invertebrates and Heterotrophic Process: 1997 Revision. ES/ER/TM-126/R2. U.S. Oak Ridge National Laboratory, Oak Ridge TN, USA. Available at http://www.hsrd.ornl.gov/ecorisk/tml26r21.pdf

Efroymson RA, Will ME, Suter GW II, et al 1997c. Toxicological Benchmarks for Screening Contaminants of Potential Concern for Effects on Terrestrial Plants: 1997 Revision. ES/ER/TM-85/R3. US Department of Energy, Washington, DC, USA. Available at http:// www.esd.ornl.gov/programs/ecorisk/documents/tni85r3.pdf

Enell A, Reichenberg F, Warfwinge P, et al 2004. A column method for determination of leaching of polycyclic aromatic hydrocarbons from aged contaminated soil. Chemosphere 54:707-15

Finster ME, Gray KA, and Binns HJ. 2004. Lead levels of edibles grown in contaminated residential soils: A field survey. Sci Total Environ 320:245-57

Frilander H and Taskinen H. 1999. Legal provisions on the protection of pregnant women at work lyoterveiset Special Issue Nr. 2:18-21. Available at http://www.td.fi/Intemet/Suomi/Tiedonvalitys/ Verkkolehdet/Tyoterveiset/

Hardison Jr DW, Ma LQ, Luongo T, et al 2004. Lead contamination in shooting range soils from abrasion of lead bullets and subsequent weathering. Sci Total Environ 328:175-83

HSDB (Hazardous Substances Data Bank). 2003. Lead compounds. National Institutes of Health, US Department of Health and Human Services, Bethesda, MA, USA. Available at http://toxnet.nlm.nih.gov/

Hyvarinen V, Solantie R, Aitamurto S, et al 1995. Suomen vesitase 1961-1990 valuma-alueittain (Regional water balances in Finland between 1961 and 1990). Vesi-ja ymparistohallinnon julkaisusarja A 220. National Board of Waters, Helsinki, Finland (in Finnish)

ISO (International Organization for Standardization). 1998. ISO 13877:1998. Soil Quality-Determination of Polynuclear Aromatic Hydrocarbons-Method Using HighPerformance Liquid Chromatography. Geneva, Switzerland

ISO. 2006. ISO 11464:2006. Soil Quality-Pretreatment of Samples for Physico-Chemical Analysis. Geneva, Switzerland

ITKC (Interstate Technology and Regulatory Council). 2003. Characterization and Remediation of Soils at Closed Small Arms Firing Ranges. Technical/Regulatory Guidelines. Small Arms Firing Team, Interstate Technology and Regulatory Council. Available at http://www.itrcweb.org

Jeltsch U and Pyy 0.1994. In:PuolanneJ, Pyy O, andjetsch U, Eds, Contaminated Soil Sites and Their Management in Finland, Contaminated Soil Site Survey and Remed

Source: Human and Ecological Risk Assessment

More News in this Category