The ecological RA will be conducted for CSSA using the most current protocols and toxicity information available at the time of the assessments. Preference will be given to reference documents obtained from the TNRCC, including the Interim Guidance for Conducting Ecological Risk Assessments at Corrective Action Sites in Texas (TNRCC, 1999). Main reference documents and applicable guidance were described in Section 1.2. The technical approach to be followed during performance of the ecological RA is outlined in this section.

Development of the initial list of COPCs for the ecological RAs will be in a manner similar to that described for the human health RAs, but will be subject to further refinement based on toxicity screening against ecological benchmarks. The ecological RAs will evaluate soils, groundwater, surface water, and sediment exposure pathways.

Similar to the human health RAs, the framework for the ecological RAs will encompass four main components. The implementation sequence of these components is illustrated in Figure 7.1. The components are:

Problem Formulation. The description of potentially exposed ecosystems; identification of representative plant and animal communities; development of a CSM that identifies affected media, exposure scenarios, pathways, and ecological receptors; and selection of appropriate assessment and measurement endpoints.

Characterization of Exposure. The evaluation of the receptor exposure conditions and development of contaminant exposure values (i.e., the C-terms) for direct contact with the affected soil, and for concentrations transferred to consumer organisms through food ingestion (i.e., as determined through the use of transfer factors).

Characterization of Ecological Effects. The selection of chemical-specific screening values for potential effects (benchmarks), and the extrapolation of these values to the site receptors for use as the measurement endpoints for the assessment.

Risk Characterization. The use of ecological hazard quotients as indicators of potential risks, and the qualitative evaluation of uncertainties associated with the ecological RA process.

7.1 - Ecological Risk Assessment Overview

This technical approach for the ecological RAs conforms to 30 TAC �350.77 and the TNRCC Interim Guidance (TNRCC, 1999) and describes a tiered analysis for screening sites and/or site contaminants for subsequent, more complex risk evaluations. The risk characterization results from each tier are used as the basis for scientific/management decisions for remedial actions and/or to determine the need for implementation of more detailed evaluations of ecological risk.

The initial step of the tiered ecological risk assessment process involves completion of a Tier 1 Exclusion Criteria Checklist, 30 TAC �350.77(b) (Appendix 1). If the results of completion of the checklist indicate that no unacceptable ecological risk exists due to incomplete or insignificant exposure pathways, then the ecological RA will be complete. However, if the affected property does not meet the Tier 1 exclusion criteria, the property will be evaluated, as needed, in accordance with subsection(s) (c) Tier 2: screening-level ecological risk assessment and/or (d) Tier 3: site-specific ecological risk assessment of 30 TAC �350.77. If, as indicated by the results of the Tier 1 Exclusion Criteria Checklist, unacceptable ecological risk exists, a Tier 2 screening-level ecological RA will be performed following the procedure outlined below.

The implementation sequence for main elements of the ecological RA process and their interactions are illustrated in Figure 7.2. The key elements are:

Development of a CSM that identifies significant exposure routes and representative receptors.

Identification of ecological COPCs on the basis of an analysis of site attributes (i.e., analysis of site operational history and background screening for inorganics), frequency of detection, and benchmark screening.

Estimates of vegetation exposures through COPC uptake from soils and wildlife exposure to COPCs through food ingestion on the basis of potential chemical transfer through the food chain and the behavior and mobility of the receptor organisms.

Calculation of ecological hazard quotients, which are chemical- and receptor-specific ratios between the calculated exposure concentration and an appropriate reference value intended to be indicative of safe threshold levels for receptor organisms.

The evaluation of risk will be conducted for each site by calculation of ecological hazard quotients for general receptor groups such as vegetation, and for representative species to be selected from the array of wildlife present at CSSA.

The use of hazard quotients for ecological risk characterization will be complemented with existing information on plant communities and wildlife as an independent line of evidence of potential ecological risk. This information will be used to compare individual sites to reference locations in terms of spatial coverage (density) and diversity of vegetation, and population abundance and diversity of species. The use of multiple lines of evidence for the ecological RA is consistent with USEPA (1997b) guidance. Available biota monitoring and field study reports for CSSA include the following:

Background Information Report (Parsons ES, 1998);

Integrated Cultural and Natural Resources Management Plan (SAIC, 1997);

Camp Stanley Storage Activity Habitat Evaluation (Engineering Science, 1992);

Annual golden-cheeked warbler and black-capped vireo surveys.

7.2 - Problem Formulation

For the ecological RA, it is anticipated that a near-surface interval and a deeper root-zone interval will be evaluated as the ecological receptor exposure intervals in soil. Site-related COPC concentrations in affected media to which ecological receptors may be exposed will be statistically summarized for each site as described for the human health risk assessment. Similar to the human health approach, maximum site concentrations will be used to perform an initial site screening. If necessary to calculate C-terms, the exposure concentrations to which ecological receptors may be exposed will be the lesser of the 95 percent UCL or the maximum detected concentration for each COPC in a given exposure interval for a specific exposure area.

As described in Section 4.3.1 and Section 4.3.2 for the human health RA, COPCs will be further refined through an analysis of frequency of detection and background screening. Basewide and/or site-specific background data obtained from previous reports will be used to help establish which chemicals can be considered to represent site-related activities.

Available reports have documented the predominant biotic communities present in the CSSA area. These documents and other available references will be used to help identify the various biotic communities located in and near CSSA.

A preliminary CSM for typical exposure of vegetation and wildlife to chemical substances in environmental media is presented in Figure 7.3. The main transfer mechanism is the uptake of chemical substances by the site vegetation and organisms, and subsequent transfer along the food chain through insects, small herbivores, and larger mammals. In turn, these organisms support multiple predators that include carnivorous mammals, birds, and snakes. The transfer pathway from insects includes an intermediate step through secondary consumers (e.g., insectivorous mammals, birds, and reptiles).

A second exposure pathway is the ingestion of groundwater from on-site wells through cattle and wildlife watering facilities supplied by CSSA. Other pathways include the direct incidental ingestion of soil, surface water, or sediment by wildlife during foraging, grooming, and/or burrowing. This type of direct exposure is accounted for in the risk calculations by incorporating soil ingestion as a component of the diet of each receptor organism. Although dermal contact and inhalation are two other possible direct mechanisms for exposure to contaminants by burrowing and/or swimming organisms, their contribution to risk is likely to be less than the contribution of other exposure routes. Currently there are no adequate data to quantify potential exposures of, or to determine potential effects on, wildlife from dermal contact or inhalation. Therefore, these exposure routes will not be quantitatively evaluated in the ecological RA, but will be addressed qualitatively in the uncertainty analysis.

The ecological RA will assume that current site conditions will be maintained for the foreseeable future. Therefore, at exposure areas where current site conditions contribute to habitat constraints that restrict the presence of ecological receptors (e.g., where structures cover the ground surface or where site components are below grade), evaluation of ecological risks may not be necessary because pathways to receptors are not complete.

A broad assessment endpoint has been adopted for the ecological RA, and where appropriate, more narrowly defined assessment endpoints (and corresponding measurement endpoints) may be developed for individual areas. The broad assessment endpoint is the determination of adverse effects induced in ecological receptors at the population level by exposure to site-related chemicals in affected media. The measurement endpoints selected to evaluate the assessment endpoint are no-observed-adverse-effect levels (NOAELs). Chemical- and receptor-specific NOAELs will be used as the reference values for calculation of hazard quotients. This is a conservative approach that is protective of site organisms because adverse effect levels might require exposure to concentrations that are much more elevated than those indicated by the NOAEL. Toxicological studies frequently use test doses that are increased in a logarithmic sequence, a procedure that often results in an order-of-magnitude gap between the no-effect and the lowest-observed-adverse-effect level (LOAEL).

For wildlife species, the ecological RA will focus on the evaluation of potential population-level effects on receptor species selected to be representative of general biotic groups (i.e., feeding guilds) present at the sites. Potential effects at the population and ecosystem levels also will be inferred if a significant risk is estimated for multiple receptors or for organisms that are key components of the food web.

Existing information for ecological habitat and biological characterizations has been used to identify potential receptors and exposure pathways for each site and for the installation in general. Existing documentation on biological communities at or adjacent to the sites will be used to identify species and their related home range sizes. Species home ranges will be used, where appropriate, to evaluate the site contamination level in reference to a species home or feeding ranges.

Any threatened or endangered species that may exist in the area will be identified and appropriate safety factors will be applied to the benchmark exposure concentrations. Two endangered bird species, the black-capped vireo and the golden-cheeked warbler, potentially occur on the facility; and several invertebrates are proposed for listing which have potential to occur in the area.

For each site at CSSA, the preliminary CSM will be evaluated and revised as necessary. Only those pathways determined to be complete will be retained for the ecological RAs. Justification for retaining or excluding exposure pathways in the ecological RA will be provided in reports.

The organisms potentially exposed to contaminants in any media are commonly referred to as ecological receptors. For direct exposure of vegetation to chemical substances in soils, the risk evaluation will be based on general reference values that were derived from combined toxicological data for several plant species.

For wildlife exposure through food ingestion, specific receptors will be selected for risk evaluation, taking into account the potential for exposure (e.g., food ingestion rate, mobility) and sensitivity to individual contaminants. The use of commonly occurring species from various trophic levels as ecological receptors is a screening method for potential effects on site organisms, and is intended to reduce the uncertainty in the risk analysis. When potential adverse effects are identified for a specific receptor, a potential risk is also assumed for other wildlife species having similar diet composition and mobility, and for endangered, threatened, or special-status species potentially present in the site vicinity. As noted in Section 7.2.3, current and future land use scenarios, and their resulting impacts on site ecology, are assumed to be essentially the same for purposes of the ecological RA.

Several wildlife species are commonly associated with plant communities dominant in the area. The species representative of various trophic levels known or expected to be present at the sites will be selected as potential receptors for the ecological RAs. Threatened or endangered species with potential to occur on the site will also be included in the RAs.

Aquatic biota may be exposed to site contaminants when surface soils are transferred to drainage channels leading to Salado, Leon, and Cibolo Creeks. As soils are deposited, sediment associated organisms come into contact with soil constituents. Potential effects on these organisms will be assessed on the basis of sediment quality guidelines that are protective of the overall aquatic community.

Vegetation, the base of the herbivore food chain, is primarily exposed to soil constituents by uptake of contaminants through the root system. Some chemical substances are largely retained in underground plant tissues while others are translocated, to a variable extent, in shoots, leaves, fruits, and seeds. For the ecological risk evaluation, the soil depths under consideration will include two depth intervals: a near-surface interval (e.g., from 0 to 0.5 foot) and a deeper, root zone interval (e.g., from 0.5 to 5 feet). These intervals follow the TRRP definitions and include the depths at which root biomass of most grasses, forbs, and small shrubs occur, as tabulated in Table 7.1 (data presented by Bell, 1992). Also, as previously discussed, the depth of the selected exposure intervals will depend on the available sampling data.

Soil microorganisms and other invertebrates are directly exposed to soil constituents. For earthworms, the exposure route is a combination of dietary intake and dermal contact. These two types of exposure are not considered independent in the ecological RA because, in practice, toxicological effect levels (benchmarks) documented in laboratory and field tests are measured in terms of soil concentrations, thus including both types of exposures. Earthworms are good candidates for risk evaluation of the contaminant transfer pathway based on detritivore organisms.

Table 7.1 - Average Maximum Root Depths of Plants

C-terms will be calculated for each COPC in each soil exposure interval as previously described, and will be used as the exposure concentrations for plants, and for the incidental soil ingestion exposure route for wildlife organisms. This approach is conservative because it assumes that soil constituents are completely bioavailable for uptake by organisms. Site-specific soil properties, however, tend to reduce the bioavailability of contaminants. For example, in the case of metals such as cadmium, copper, lead, mercury, and zinc, their uptake by vegetation and their potential phytotoxicity have been related to multiple soil geochemical parameters, including mineralogy, pH, redox potential, and organic material content (EPA, 1983). Furthermore, exposure assessments for metals often assume that the metal is present in its most toxic form, typically the inorganic salt form evaluated in many toxicological tests. There are multiple initiatives underway to calculate actual availability of soil contaminants to vegetation and soil organisms (i.e., Alexander et al., 1995). Results of a study by Hansen and Chaney (1984), shown in Table 7.2 illustrate potential limitations in the transfer of metals and other inorganic substances from contaminated soils to plants and herbivores. These geochemical constraints on bioavailability will be noted in the uncertainty analysis of the Screening Risk Assessment/Risk Assessment (SRA/RA) reports.

Aquatic biota might be exposed to site COPCs by the transfer of surface soils to surface water by runoff from rainfall events. If a significant transfer pathway for chemical substances is present from site soils to aquatic organisms, this exposure pathway will be evaluated qualitatively in the ecological RA. Temporal ponding of water, that may occur in some locations, will not be considered suitable for development of a complex aquatic ecosystem; however, short-term exposures will be qualitatively evaluated on a case-by-case basis if needed.

Any qualitative estimates of contaminant concentrations in surface water and/or sediment due to overland flow will be based on a simplistic partitioning model. These estimates will be compared to state and federal surface water quality and sediment criteria to provide a qualitative evaluation of potential adverse effects to aquatic species.

7.2.9.1   Direct Soil Ingestion

For many wildlife species, diets may include the incidental ingestion of soil. Typically, the ingestion rate is less than 10 percent of the diet for most wildlife species. The exposure to contaminants by direct soil ingestion may be significant for herbivores feeding on underground portions of plants and for insectivores feeding on soil dwelling invertebrates (EPA, 1993b). Table 7.3 lists example soil ingestion rates for receptors that may be among those selected for evaluation in the RA.

7.2.9.2 Food Intake

For terrestrial wildlife, dietary exposure through consumption of vegetation is considered the primary exposure route for chemical substances. The transfer of chemical substances from the site vegetation to wildlife species follows three main sub-paths.

Consumption of vegetation by insects and other small arthropods such as grasshoppers, ants, and beetles. These organisms support insectivore organisms such as lizards and various songbirds, which in turn are consumed by predators such as snakes and birds of prey.

Consumption of vegetation by small rodents and seed-eating birds. These organisms serve as the transfer mechanism to multiple predators that include snakes, birds of prey, the coyote, and the fox.

Consumption of vegetation by large grazers such as deer. These grazers have relatively few predators within CSSA boundaries with the possible exception of the coyote. In practice, the main transfer mechanism for this sub-path is consumption by carrion eaters.

7.2.9.3   Dietary Intake Calculation

Exposure to a chemical substance by dietary intake will be determined by:

1. the chemical concentration in the food source;

2. the food ingestion rate (IR) of the consumer organism; and

3. the area use factor (AF), which accounts for the fraction of the diet that an organism actually obtains from the site.

The AF takes into consideration the dietary fraction derived from a site based on the organism�s foraging/feeding range (i.e., the mobility factor, which is the ratio between the site surface area and the foraging/feeding range), and seasonal exposure that limits exposure to certain periods of the year. High mobility animals, such as raptors, have extensive foraging ranges and are known to obtain their diet from multiple locations, while low-mobility organisms such as invertebrates and small wildlife have a higher degree of exposure because most or all of their diet is derived from a particular site.

Table 7.3 - Example Wildlife Exposure Parameters for Ecological Risk Assessment

Based on the three above listed exposure parameters, the dietary exposure for an organism can be calculated as:

Dietary Exposure = Concentration in Diet x IR x AF

Where:

Dietary exposure is expressed as a daily dose per unit body weight (mg/kg/day);

Concentration in diet is given in mg/kg (dry weight basis);

IR is the daily ingestion rate of food and/or water consumption, expressed as a fraction of the organism�s weight (kg food or water/kg body weight/day); and

AF is the area use factor, a unitless value that accounts for the organism�s foraging range (hectare/hectare) and/or seasonal presence (fraction of the year).

As summarized by USEPA (1993b), Table 7.3 shows available data on foraging ranges and soil, food, and water intake rates for representative ecological receptors present. The data presented will be updated with more current and complete information as appropriate before being used in the ecological RA. Values proposed for use in the RA will be the midpoints of the reported foraging ranges and the reported ingestion rate data. Table 7.3 also summarizes the predominant diet composition of the example receptor organisms.

When the concentration of a chemical compound in various types of foods is known or estimated, its dietary concentration can be calculated based on the relative contribution of each type of food to the organism�s diet. The nomenclature used for identification of the concentration of chemical substances in five major food sources, and their relative contribution to an organism�s diet (as "f_I" the fraction in the total food ingested) are tabulated below:

For a species whose diet includes multiple foods sources, the dietary concentration of a chemical substance is calculated as:

Concentration in Diet = f_SC_S + f_VC_V + f_I C_I + f_HC_H + f_CC_C

Where:

f_S + f_V + f_H+ f_I + f_C = 1

For species with diets composed predominately of only one or two food sources, the contributions of other food sources become negligible. For predatory organisms such as raptors that rely on herbivores and small carnivores as their main food source (fS = fV = fI = 0), the dietary concentration simplifies to:

Concentration in Predator�s Diet = f_HC_H + f_CC_C

Because organism tissues are not routinely monitored during typical site investigative studies (e.g., RFIs) or even during initial phases of ecological RA studies, the concentrations of chemical substances in wildlife diets are often unknown. In such cases, it is necessary to estimate tissue values from concentrations measured in soils. Tissue values will be estimated from measured concentrations in soils when analytical data are not available, or are not applicable to the sites under consideration.

To evaluate potential effects on wildlife by food ingestion, the concentration in the diet can be estimated from concentrations detected in soils using unitless transfer factors (TFs). These TFs are empirical values that account for the potential accumulation of a chemical in organism tissues at each transfer step in the food chain.

Based on the use of TFs, the concentration of a chemical substance in tissues of four types of organisms can be estimated from the soil concentration (CS) as follows:

Concentration in Vegetation Tissues: C_V = C_S x TF_V

Concentration in Invertebrate Tissues: C_I = C_S x TF_I

Concentration in Herbivore Tissues: C_H = C_V x TF_H

C_H = C_S x TF_V x TF_H

Concentration in Small Carnivore Tissues: C_C = C_I x TF_C

C_C = C_S x TF_C x TF_I

For an omnivorous organism having food derived from multiple food sources, the concentration of a chemical substance in its tissues (CO) can be estimated as:

C_O = f_SC_S + f_VC_V + f_I C_I + f_HC_H + f_CC_C

Replacing the concentration in diet as a function of soil concentration using dietary transfer factors results in:

C_O = f_SC_S + f_VTF_VC_S + f_ITF_IC_S + f_HC_S (TF_H x TF_V) + f_CC_S (TF_C x TF_I)

And finally, rearranging the equation as a function of the soil concentration (CS), the concentration in the tissues of the omnivore organism is obtained as:

C_O = C_S [ f_S + f_VTF_V + f_ITF_I + f_H(TF_H x TF_V) + f_C(TF_C x TF_I) ]

Transfer factors used in exposure-rate calculations for organic compounds can be obtained from the technical literature or can be calculated using empirical equations relating potential bioaccumulation to the tendency of the substance to accumulate in lipid (fat) tissues (as measured by the octanol-water partition coefficient, Kow). For example, a literature value of 0.2 is available for the soil-to-vegetation TF for the organic compound benzo(a)pyrene, which has a log K_ow of 8.1 (Bell, 1992). Table 7.4 presents example literature-derived TFs for the soil-to-vegetation (TF_V) and vertebrates to predators (TF_c) pathways for metals. A default value of 1.0 is used in the table when estimates of the TFs were unavailable or when the potential for bioaccumulation of a given metal is known to be low. This default value conservatively indicates that concentrations of a chemical in the plant or consumer organism tissues are similar to those found in the soil or food source, respectively.

7.2.10.1   Transfer from Soil to Vegetation Tissues

Transfer of organic substances from soil to vegetation tissues can be calculated from the K_ow using the following empirical equations for roots and foliage:

log (TFV for net root uptake) = 2.0 - 0.11 log K_ow

log (TFV for foliage) = 1.588 - 0.578 log K_ow

Net root uptake is based on experimental data relating the uptake of multiple pesticides by barley roots to their K_ow (Table 3.6 of Bell, 1992). For foliage, a geometric correlation equation is available from data on bioconcentration of 29 chemicals in cattle-feed grasses as a function of K_ow (Travis and Arms, 1988).

The example TFs from soil to vegetation (TF_V) presented for metals in Table 7.4 were calculated from concentration ranges in plant tissues, as summarized by Baes, et al. (1984). The TF was obtained by dividing the midpoint of the reported ranges for vegetative tissues by the average concentration of the element in natural soils. Data for non-vegetative tissues (fruits and tubers), which have lower levels of metals bioconcentration than vegetative tissues (USDOE, 1998a; USDOE, 1998b; USDOE, 1998c), were not used to calculate the presented transfer factors.

7.2.10.2   Transfer to Vertebrate Tissues

For the purposes of demonstration in this technical approach document, a single transfer factor (TF_H = TF_C) was used in Table 7.4 to account for accumulation of metals from plants/prey organisms (invertebrates, small vertebrates) to their consumer organisms. A default value of 1.0 was used for metals, with the exception of mercury, which is known to bioaccumulate in high trophic levels of the food chain (EPA, 1993c). For organic substances, the transfer factors TF_H and TF_C will be calculated as a function of the K_ow using literature data (e.g., Howard, 1990) to represent the expected increase in tissue concentrations from a secondary consumer (e.g., a food chain multiplier for trophic level 3 consumers) to a higher order consumer (e.g., trophic level 4). The most current data available will be used to calculate TFs once COPCs have been determined and receptor selection has been finalized.

7.3 - Characterization of Ecological Effects

The initial screening phase of the ecological RA will involve the comparison of COPC concentrations in affected media to screening values that identify threshold levels below which adverse effects on receptor organisms are unlikely. These screening values for potential effects on site organisms are referred to as benchmarks. For evaluation of effects on organisms by direct exposure, the value for comparison to benchmark is the concentration of a chemical substance in the medium. For evaluation of exposure through food and/or water ingestion, benchmarks for specific chemical substances are dietary doses associated with various adverse levels documented under experimental conditions.

Currently, there are no universally-accepted methods for deriving benchmarks used in ecological RAs. For different locations and exposure scenarios, multiple approaches and benchmark values have been used in the evaluation of ecological risk. Commonly used benchmarks document toxicological endpoints for individual species and, less frequently, potential effects on several test species or assemblages of organisms. The following section describes the proposed benchmark derivation methods and rationale for their use in the initial screening assessment of potential ecological risk.

Table 7.4 - Example Transfer Factors for Potential Bioaccumulation of Metals in the Terrestrial Food Chain

Benchmarks for evaluation of potential effects of contaminants on terrestrial vegetation will be obtained from various sources, including the most current screening reference values used by the Environmental Sciences Division of the Oak Ridge National Laboratory (ORNL) to evaluate the need for remediation at hazardous waste sites within the jurisdiction of the U.S. Department of Energy (Will and Suter, 1995a and 1995b). These screening benchmarks identify soil concentrations with low potential for effects on vegetation based on toxicological data for multiple test organisms. In the evaluation of potential effects of COPCs in soils on terrestrial wildlife, benchmarks likely will be based on toxicological data for individual test species, such as those summarized by Sample, et al. (1996). Aquatic systems will be based on USEPA and state criteria, Suter and Tsao (1996), and Jones, et al. (1996).

Due to limitations of the toxicological database, many benchmarks are commonly obtained by extrapolation of acute-effect levels (such as the LC₅₀, the concentration that is lethal to 50 percent of the test organisms) to a no-adverse-effect level, the measurement endpoint selected for the risk evaluation. Extrapolation is also often required between related organisms or chemical groups when toxicity information is unavailable for the target receptor or the COPC. The basis for the extrapolation methodology is described in Section 7.3.3.

Benchmarks for screening of potential effects of metals on terrestrial plants were developed by Will and Suter (1995a) based on a summary of reported low-effect levels on vegetation growth and production. Will and Suter ranked reported data and used the 10th percentile of the data distribution as a screening benchmark intended to provide a 90 percent level of protection for plant communities. When phytotoxicity data were insufficient for calculation of a 10th percentile, the lowest-effect level reported was used as the benchmark value. In some cases, low-effect concentrations were obtained by extrapolation from lethal effects data (low-effect level = 20% LC₅₀).

Benchmark values for organic substances are obtained primarily by extrapolation of acute-effects levels for a single species (an EC₅₀) to a no-effect level. For pesticides, benchmarks usually represent the maximum application rates resulting in no-observed-effect levels (NOELs) on crops, as reported by the USEPA (1983). These values are extrapolated to soil concentrations assuming a 6-inch penetration in the soil, and negligible pesticide degradation or accumulation. When unavailable, the NOEL was obtained by extrapolation of lowest-observed-effect levels (NOEL = 10% LOEL).

Benchmarks for screening of potential effects of metals on mammals and birds by dietary ingestion represent the NOAELs for dietary intake reported for a selected test species. In the selection of final benchmarks for the RA, preference will be given to toxicological data documenting effects on reproduction and development associated with chronic exposures. Most NOAELs used in risk evaluations, such as those summarized by Sample, et al. (1996), are data obtained using laboratory mice, rats, and mallard ducks as test organisms. Other toxicological data for benchmark development are available in other summary documents by USEPA (1993b), Extoxnet (1993), Hill and Camardese (1986), Hudson, et al. (1984), Sax (1986), United States Fish and Wildlife Service (USFWS) (1986a and 1986b) and Will and Suter (1995b). Final wildlife benchmarks will be developed from most current available information once receptor selection and COPC determination are finalized.

Due to limitations inherent in the ecotoxicological database, many benchmarks must be obtained by extrapolation of acute lethal effect levels to NOAELs, the measurement endpoint selected for the risk evaluation. Extrapolation is also required among related organisms or chemicals when toxicity information is unavailable for the target receptor or the specific COPC.

7.3.3.1 Endpoint Extrapolations

For chemical compounds without a reported chronic NOAEL, extrapolations may be made from subchronic tests, and from tests documenting LOAELs or acute effect levels such as the dose lethal to 50 percent of the test organisms (LD50). Extrapolation factors are as follows:

Chronic NOAEL = 10% Chronic LOAEL

= 5% Subchronic LOAEL

= 1% LD50

7.3.3.2   Site-Specific Extrapolations

Appropriate toxicity data are often unavailable for a site-specific COPC or a specific ecological receptor because such data are limited to relatively few chemical compounds tested with laboratory organisms and plant species. For the ecological RAs, NOAELs will be extrapolated, if necessary, between chemical compounds, and from test organisms to the selected ecological receptors, as described below.

Extrapolation Between Chemicals. When toxicity data for a specific COPC are not available, the value will be obtained from a surrogate compound. When available, data from a closely-related chemical will be used to generate the toxicity reference value (TRV). In cases where this information is unavailable, the compound with the lowest toxicity concentration value reported within a given chemical group will be used as a surrogate compound (i.e., the most conservative value). Surrogate compounds may be used for COPCs in the following chemical groups: VOCs, polycyclic aromatic hydrocarbons (PAHs), chlorinated benzenes, insecticides, and herbicides.

Extrapolation Between Different Species. Toxicity data for dietary intake by terrestrial wildlife is largely based on laboratory testing of selected organisms. Among mammals, the most frequently tested organisms are the laboratory rat and mouse. Commonly tested bird species include the mallard duck and various quail species.

Following USEPA�s approach for deriving human toxicity values from animal data, experimentally derived NOAELs and LOAELs are used to estimate NOAELs for wildlife by adjusting the dose according to differences in body size. Based on a data review for mammal species by Sample, et al. (1996), this relationship can be best expressed using a scaling factor of the form:

Scaling factor = (weight of wildlife mammal species / weight of test species) ^0.25

Unlike mammals, most reported scaling factors for bird species do not significantly differ from a value of 1 (Sample, et al., 1996). Accordingly, a scaling factor of 1 is considered appropriate for interspecies extrapolation among birds.

7.4 - Ecological Risk Characterization

The assessment of ecological risk will be based on the calculation of ecological hazard quotients (HQs). This parameter is a semi-quantitative estimate of possible risk that is calculated as the ratio between exposure concentration for a given chemical substance and an applicable reference toxicity value, the exceedance of which identifies possible adverse effect levels on a receptor. For the RAs, it is proposed that the measurement endpoint, the NOAEL, will serve as the reference value. Because of the numerous conservative assumptions typical of ecological RAs, and several resulting uncertainties associated with HQ calculation, HQs only provide order-of-magnitude estimates of the potential for adverse effects, not exact measurements of actual effects on receptor organisms. For the ecological RAs, the characterization of potential effects of chemical substances on selected receptor organisms will be based on the following guidelines proposed by Barnthouse et al. (1986) and adopted in subsequent evaluations of ecological risk (EPA, 1989d; Watkin and Stelljes, 1993; Menzie et al., 1993):

Adverse effects are not expected for HQ values equal to, or less than, one;

A potential for environmental effects is indicated by HQ values greater than one.

Where appropriate, the HQs for COPCs with similar modes of action may be summed by receptor for each SWMU. The resulting cumulative ( HQ is referred to as an ecological hazard index (HI).

Similar to the human health RA, a qualitative analysis will be made of the uncertainties associated with the quantification of ecological risk. This analysis will be based on the potential of underestimating or overestimating the risk of adverse effects in organisms using four uncertainty categories: low, moderate, high, and unknown potential.

In general, assumptions that will be used in the evaluation of receptor exposures and potential adverse effects are conservative and tend to overestimate actual risk values. Extrapolation factors and TFs for potential bioaccumulation of chemical substances will be selected to provide an elevated level of protection for sensitive ecological receptors. Specific parameters and assumptions that affect the uncertainty of the risk characterization will be noted in the uncertainty analysis. Such parameters/assumptions may include assumptions about the bioavailability of COPCs in soils, the exposure factors applied (e.g., foraging ranges and seasonality of exposure), and data limitations (e.g., depths of sample collection).

[Next Section]