PUBLIC HEALTH ASSESSMENT
AIR AND GROUNDWATER PATHWAYS ANALYSES
AGANA POWER PLANT
September 22, 2004
- Exposure to contaminants from eating foods raised locally.
- Breathing in contaminants—PCBs that might evaporate from the Agana Swamp.
- Breathing in contaminants—dusts from roads near the former APP.
- Breathing in contaminants—smoke from fires in the Agana Swamp.
- Unearthing buried wastes from various sources.
- Past occupational exposures.
- Exposure to chlordane in the Mongmong area from soil or water.
| How Do Animals Come Into Contact With Contaminants? |
Due to their extensive use and persistence, many contaminants (e.g., PCBs, dioxins, and some inorganic compounds such as metals) are widespread throughout the environment. Studies have shown that free-range animals can ingest substantial amounts of soil while feeding. Many free-range animals (e.g., chickens) need a certain amount of soil and grit in their diet to assist in digestion. If these animals are feeding on contaminated land, they could be ingesting persistent chemicals that bind to soil particles and accumulating these chemicals in their tissues (Beyer et al. 1994).
Free-range animals may also be exposed to contaminants via ingested vegetation and water, as well as other food sources (e.g., insects, earthworms, and other small animals). The overall contribution of these other potential sources of exposure varies depending on what type of animal is being evaluated and the nature and extent of the contamination.
Some people on Guam harvest or catch animals or use locally obtained animal products to supplement their typical diets. These animals include free-range chickens, which are quite prevalent on the island, feral or wild swine (pigs), sheep, and perhaps a small number of free-range steers and milk cows that may be used primarily for meat or for milk or other dairy products. A number of individuals mentioned that they would like to obtain and raise carabao and suggested that this does occur on occasion. In addition to catching and eating fish, turtles, and apple snails from the swamp and river, some people may also ATSDR has received information from residents in the Mongmong area that certain species of turtle are harvested and consumed from the Agana Swamp. Turtles can accumulate higher levels of PCBs and pesticides. However, GEPA and Navy contractors have not been able to find turtles in the swamp to sample. ATSDR recommends that anyone with knowledge about where turtles are commonly harvested in the Agana Swamp contact GEPA or the Guam Government's Department of Agriculture.harvest land snails (e.g., the giant African snail) that can be found in many residential areas on Guam. Although Guam residents occasionally consume all of these animals, ATSDR has not identified any populations in the Mongmong area that subsist primarily on locally obtained animals. It is ATSDR's current understanding that the vast majority of the villagers on Guam do not eat land snails, although some villagers do consume apple snails.
How Would Free-Range Animals Most Likely Be Exposed to Contaminants Near APP?
Mostly low levels of PCBs and dioxins have been detected in surface water, sediment, and soils around the Mongmong area near APP. The concentrations detected in soil and water do not pose a public health hazard for people who come into contact with these contaminants in their daily activities (walking, swimming, gardening, etc.). Some residents, though, have asked whether free-range animals used for food could accumulate harmful levels of PCBs from the relatively low concentrations detected in the environment near APP.
Wildlife may be exposed to organic chemicals such as PCBs and dioxins through ingesting contaminated soils, vegetation, drinking water, or the foods they eat. For those organic chemicals that bioaccumulate up the food chain, people can be exposed by eating local wildlife that graze in contaminated areas.
ATSDR has evaluated this potential exposure pathway by concentrating on soil ingestion by free-range animals. Although there are other routes of exposure for animals, it is likely that the animals of interest in the Mongmong area would accumulate PCBs in their bodies primarily from ingesting soil or very small organisms (e.g., earthworms) that contain soil with PCBs or that may have accumulated PCBs. The local wildlife that are most likely to be consumed by people from villages on Guam are mostly those that eat small animals or insects. The invertebrates sampled near APP included apple snails and worms. The concentrations of PCBs in samples from these invertebrates were either very low or not detected (U.S. Navy 2003).
It is unlikely that the vegetation will be a significant source of PCBs since plants do not readily accumulate organic chemicals into their root systems. PCBs are commonly detected at trace levels (well below levels that can cause health effects) in fruits and vegetables that are selected for the U.S. Food and Drug Administration's market basket studies. PCBs can be deposited onto the surface of vegetation and ingested. PCBs were not detected in the plant tissue and taro samples collected by the Navy, and it is unlikely that this is a significant exposure pathway. PCBs in surface water samples collected by the Navy were only detected in a few samples detected at very low concentrations (U.S. Navy 2003).
Where Were the Highest PCB Concentrations in Soil Detected?
PCBs and very low concentrations of dioxins and furans (i.e., polychlorinated dibenzo-p-dioxins [PCDDs] and polychlorinated dibenzofurans [PCDFs]) have been detected in surface soil, both on APP property and offsite near the APP. Soil, sediments, and surface water have been sampled by the U.S. Navy. Initial sampling was based on known locations of PCB-containing equipment, visible soil staining, the direction of runoff flow, and interviews with previous APP workers. The Navy also selected residential areas in close proximity to APP and conducted PCB sampling on private properties (ATSDR 2003).
The Navy has collected many soil samples, both at the surface (0 to 6 inches bgs) and below the surface (1 to 50 feet bgs). The levels of PCBs in the soil below the surface are not very different from the levels of PCBs detected near the surface (see Table 2 below) (U.S. Navy 2003). The samples collected were analyzed for some combination of PCBs, dioxins and furans, PAHs, VOCs, SVOCs, and metals. ATSDR reviewed sampling data from the Navy. Only PCBs were detected in surface soil at levels that warranted further evaluation by ATSDR.
As Table 2 shows, the highest PCB concentrations (7,800 ppm) detected in surface soil, either onsite at APP or offsite, was at the Agana Springs (Upland) sampling location. This location is approximately 3/4 mile southwest of APP and was initially selected as a reference location by the Navy. The PCBs detected at this sampling location were unrelated to activities at APP. Additional investigations revealed that this location was an old transformer pad and a pumping station used by the Navy (Mathew McNeff, Earth Tech, personal communication, April 17, 2003). The maximum concentration of 7,800 ppm is more than 4 orders of magnitude larger than EPA's risk-based concentration (RBC) for PCBs in residential soil. But this high value was only detected in one sample, and it is not representative of other samples detected in the general area: for example, the second-highest PCB concentration detected in the area was 7 ppm (U.S. Navy 2003).
PCBs were also detected in samples collected at the APP Drainage Outfall (Upland area) (maximum 25 ppm) and APP West Drainage Area (maximum 29 ppm). PCBs were either detected at very low concentrations or were below the method detection limit in soil samples collected from the residential locations. The maximum PCB concentration detected at the residential properties was 0.6 ppm (Aroclor 1254). PCB concentrations detected in samples collected on APP property (maximum 99 ppm) were somewhat higher than concentrations detected in the offsite Drainage Outfall (Upland area) and APP West Drainage areas
The Navy has evaluated areas likely to be impacted by APP operations. For other areas in the Agana Swamp and surrounding areas, raw sewage, septic tank releases, sewage amended soils, and the use of fertilizer may result in the potential uptake of chemicals. For all locations, however, the potential uptake of heavy metals in the internal organs of livestock—especially sheep and goats—cannot be evaluated with available information.
What Types of Animals Are Most Likely To Be Obtained Locally on the Island and How Often Are They Consumed?
The island of Guam contains several types of edible terrestrial (land) animals that may be exposed to contaminants in soils or sediment. ATSDR has gathered information about the most common types of animals that are consumed and has collected available information on their potential for bioaccumulation and frequency of consumption. ATSDR has contacted the Department of Agriculture (specifically, its Agricultural Development Service and Division of Aquatic and Wildlife Resources, or DAWR), the University of Guam, and other government of Guam agencies to identify reliable information regarding villagers' consumption patterns regarding edible animals on the island, in particular in the Mongmong area. Published wildlife consumption surveys or reliable anecdotal information is very limited and in most cases was not available for the edible animals evaluated below.
Potential for Bioaccumulation
Bioaccumulants: Substances that increase in concentration in living organisms as they take in contaminated air, water, or food because the substances are very slowly metabolized or excreted.
Bioconcentration: The accumulation of a chemical in tissues of a fish or other organism to levels greater than in the surrounding medium.
Biological Magnification: Refers to the process whereby certain substances such as pesticides or heavy metals move up the food chain, work their way into rivers or lakes, and are eaten by aquatic organisms such as fish, which in turn are eaten by large birds, animals or humans. The substances become concentrated in tissues or internal organs as they move up the chain.
Source: EPA Terms of the Environment http://www.epa.gov/OCEPAterms/bterms.html
Land snails are invertebrates within the mollusk family that can be found in gardens or other vegetated habitats. Studies have shown that snails and other mollusks can accumulate metals and other environmental contaminants when they ingest soil (Gomot-De Vaufleury and Pihan 2002). Most of the studies on uptake of contaminants in snails have been associated with metals such as chromium, zinc, and lead. Most of the studies looking at the uptake of organic chemicals in snails indicate relatively low potential for bioaccumulation. For example, one study evaluated the uptake, excretion, and metabolism of phorate, a persistent organophosphate pesticide, in a species of land snail (Helix aspersa). The snails were fed a solution of 100 ppm of phorate, and the pesticide was readily absorbed within 24 hours following ingestion (Salama and Radwan 1995). However, the study showed that the phorate did not readily accumulate in the snail tissues. Instead, the phorate and/or its metabolites were rapidly excreted soon after ingestion. Another study of organochlorine insecticide residues (e.g., DDT, aldrin, heptachlor) in soils and invertebrates indicated that snails did not accumulate insecticides as readily as other invertebrates such as earthworms and slugs (Gish 1970).
The Navy has not sampled land snails near APP, but it did sample apple snails from the Agana Swamp and Agana River. Snails from the Agana Swamp would most likely be exposed to PCBs and other contaminants from sediments and organic matter ingested at the bottom or near the edges of the swamp. The apple snails that were sampled contained the lowest PCB and dioxin concentrations of all the edible food items sampled from the Agana Swamp and Agana River (see Table 3) (U.S. Navy 2003). This is consistent with the literature showing that snails do not readily accumulate organic compounds in their tissues.
Patterns of ConsumptionAlthough there are some edible species of land snails on Guam, ATSDR has not identified any published or unpublished surveys measuring their frequency of consumption there. Other than limited anecdotal information, it is not known if any of the villagers routinely harvest land snails for consumption. According to Guam's Department of Agriculture, the only edible species of land snail on Guam is the giant African snail and most villagers will not eat them. The African snail can act as a vector of human disease for eosinophilic meningitis (see ATSDR glossary for more information). This disease is caused by a parasite, which is passed to humans when they eat raw or improperly cooked snails (Tino Aguon, DAWR, Guam Department of Agriculture, personal communication, April 7, 2003).
Estimating Human Exposure From Consuming Land Snails Near APPThe public health and government of Guam officials who ATSDR contacted stated that land snails were not commonly harvested as a source of food for villagers on Guam. Even if land snails near APP were consumed by a small number of people, organic chemicals (e.g., PCBs and dioxins) would not accumulate in the tissues of the snails at levels that would pose a health concern. This conclusion is based on the relatively low PCB concentrations detected offsite, the very low PCB concentrations detected in the aquatic apple snails in the Agana Swamp, and research showing the low bioaccumulation potential of organic chemicals in snails. ATSDR concludes that consumption of land snails in the Mongmong area poses no public health hazard.
Free-Range Animals (e.g., Chickens, Swine, Cattle)
Potential for BioaccumulationThe consumption of contaminated feed (either manufactured or obtained through grazing in contaminated areas) and soil are the primary pathways of animals' exposure to environmental contaminants (Fries 1995; 1996). As previously noted, ATSDR assumes that most wildlife exposure to contaminants from APP would be from soil ingestion rather than contaminated feed. Additional assumptions about the bioavailability and bioconcentration of contaminants in free-range animals are provided in Appendix D.
Research has shown that the levels and patterns of organic chemicals (e.g., PCBs, PCDDs, and PCDFs) in tissues and eggs of chickens and other free-range animals are closely related to the levels and patterns in the soils in which they graze (Stephens et al. 1990). Chickens, cows, and swine (pigs) typically consume a substantial amount of soil during feeding. The concentration of contaminants in soil is an important factor for estimating the bioaccumulation of chemicals in animals (Eduljee and Gair 1996).
Most commercial poultry and swine production is conducted in confined operations, limiting contaminated soil as a potential source of exposure. Free-range chickens and swine, however, will have access to pasture and may be exposed to contaminated soils (Fries 1995).
In general, soil ingestion is related inversely to the availability of forage (i.e., food) for animals that rely almost exclusively on pasture land for food. Soil ingestion by free-range animals will vary considerably depending on the season (e.g., the grass-growing season versus the dormant season) or climatic conditions (e.g., wet periods versus droughts). Dairy cows and beef cattle often subsist on diets that are primarily forage, i.e., grass and other vegetative cover). Unlike swine, cattle and sheep need large land areas to graze. Estimates of soil ingestion rates range from as low as 1% or 2% of dry matter intake for cows in spring, when grass is abundant, to as high as 18% of dry matter intake during the winter, when vegetation is sparse (Fries 1995; Beyer 1994; EPA 1998a).
Patterns of ConsumptionAccording to the U.S. Department of Agriculture, data are only collected on domestically raised livestock (e.g., livestock maintained on farms) and surveys do not include any estimates of the numbers of wild or free-range animals, or of how frequently villagers consume them (Roger Beinhart, USDA Guam Agricultural Census, personal communication, April 9, 2003). This lack of information is confirmed by communications with several other government of Guam agencies. A representative from Guam Public Health's Environmental Health Branch indicated that there are few regulations or restrictions governing free-range chickens on the island and there was no way to accurately assess how often free-range chickens are being consumed (Robert Contreras, Health Inspector, Guam Public Health, personal communication, April 7, 2003).
Representatives from DAWR do not believe much hunting occurs in the Agana area: the population is dense and firearm regulations require that hunting take place at least 100 yards from a residence. According to DAWR, only a few people (two or three) raise swine and perhaps four or five people raise chickens within 1 mile of APP (Tino Aguon, Department of Agriculture, DAWR, personal communication, April 11, 2003; Gerry Davis, Acting Chief, DAWR, personal communication, May 19, 2003). A representative from the village of Agana's (Hagatna's) mayor's office noted that a typical Guam resident may occasionally eat wild chicken (e.g., once a month). Wild swine are also popular, particularly during the fiestas that often occur on the island. People typically raise these animals in their yards, however, and it is not known how often free-range animals are caught and used for food (Carmen Guzman, Agana Mayors Office, personal communication, May 28, 2003).
According to some of the local residents, very small numbers of grazing cows and sheep can be found on some parts of Guam. Rabbits are raised by some residents on the island, but they are generally caged and would not come in contact with PCB-contaminated soils; they will not be discussed further in this exposure evaluation.
Estimating Exposure From Consuming Free-Range Animals Near APPAfter compiling available site-specific information about the presence of free-range animals in the Mongmong area, as well as qualitative information about the frequency of wildlife consumption on Guam, ATSDR selected three food consumption scenarios at three sampling locations that would most likely contribute to an individual's lifetime PCB exposure. The animals or animal products considered in this exposure evaluation were free-range chickens, chicken eggs, and swine.
ATSDR conducted an extensive literature search related to the intake of organic contaminants by free-range animals and reviewed common methods used to estimate the concentrations in the edible meat of the animals. Biotransfer factors (BTFs) were used to quantify the relationship between the concentration in the animal and the concentration in the medium (e.g., soil) to which the animal was exposed. The results of the three exposure scenarios are discussed below.
Exposure Evaluation for Eating Free-Range Animals and Animal Products From the Mongmong Area
To estimate an individual's potential for exposure from consuming free-range animals in the Mongmong area, ATSDR first evaluated the nature and extent of contamination by reviewing surface water, sediment, and soil data collected by the U.S. Navy at seven locations near APP. ATSDR compared the results with corresponding screening values (i.e., our comparison values, or CVs). Three of the seven sampling locations contained PCBs in surface soil and sediment above ATSDR's screening level (see the text box). ATSDR used PCB concentrations detected in surface soil and applied chemical-specific BTFs to estimate the concentration in the animal or animal product (i.e., eggs).
PCBs Were Detected in Three Offsite Sampling Locations Near APP at Concentrations Above ATSDR's Screening Level
- Aroclor 1260 detected in surface soil
Maximum concentration: 7,800 ppm at an old transformer pad
- Aroclor 1260 detected in sediment
Maximum concentration: 7.2 ppm
- Aroclor 1254 detected in surface soil
Maximum concentration: 29 ppm
- Aroclor 1260 detected in surface soil
Maximum concentration: 25 ppm
ATSDR's comparison value for PCBs in soil is 10 ppm. This represents the Chronic Adult Environmental Media Evaluation Guide (see Appendix C for additional explanation) for Aroclor 1254.
Note: PCB removal actions at the APP West Drainage Area and Drainage Outfall (Upland) have taken place and current PCB levels in soil are expected to be below ATSDR's public health action levels. .
The levels shown in the text box represent maximum concentrations detected since sampling at APP was initiated. In February 1999, the Navy removed PCB-contaminated soil from the vicinity of the APP West drainage Area, to a depth of 2 feet below grade surface and replaced it with clean fill material (US Navy 2000). After remediation the PCB contaminant level near the church averaged 0.53 mg/kg. In April, 2000, the Navy also initiated a removal action at APP's Drainage Outfall (Upland) Area. ATSDR's evaluation of public health hazards for this location is based on the assumption that people were exposed to the average concentrations detected since sampling began. Therefore, ATSDR's exposure evaluation likely overestimates the extent of contamination in the APP West Drainage Area and Drainage Outfall (Upland)
ATSDR has no reason to expect that PCBs in animals (cattle, swine, chickens, goats, sheep, and rabbits) consuming soil vegetation or store-bought feed contaminated with PCBs at or below current tolerance levels will accumulate in the animals' fat greater than levels found in dietary bread basket studies of New Zealand, Europe and the United States.
Because of the continued reduction in PCB use around the world—including Guam—the average dietary intake of PCBs is continually being reduced. For example, the estimated dietary intake of PCBs by adults in the United States in 1978 was 0.027 g/kg of body weight per day. From 1982 through 1984, this level was reduced to 0.0005 g/kg of body weight per day (ATSDR 2000c). Today, PCB blood serum levels have been reduced for non-occupationally exposed individuals who do not consume fish to between 0.9 and 15 ppb in the studies conducted of the U.S. population. In 1986, it was estimated that 95% of Americans had PCB blood serum levels below 20 ppb, with an average of 4 to 8 ppb. In 1999, the average PCB blood serum level in a sample of consumers of sport-caught fish in the Great Lakes was 4.8 ppb (males) and 2.1 ppb (females). Infrequent consumers of fish, however, had PCB blood serum levels of 1.5 ppb (males) and 0.9 ppb (females).
To estimate an individual's dose of PCBs from consuming free-range animals, ATSDR made several health-protective assumptions about the exposure. These assumptions represent worst-case scenarios and likely overestimate the amount of contaminants to which people are exposed. The methodology and assumptions used by ATSDR are presented in Appendix D.
The Navy did not include dioxins in its list of compounds to be tested in all offsite soil samples because its analysis of dioxins in samples collected onsite at APP showed that concentrations were all below EPA's dioxin soil cleanup level of 1 ppb, which ATSDR also uses as a health guidance value. Therefore, ATSDR did not calculate TEQ (i.e., toxic equivalent) concentrations for dioxins/furans because the Navy's offsite sampling protocol did not include routine analysis of dioxins (only one soil sample collected offsite was analyzed for dioxins). The maximum 2,3,7,8-TCDD TEQ calculated in samples collected on site was 0.095 ppb (Earth Tech 2002).
ATSDR's Estimated PCB Concentrations in Free-Range Animals and Animal Products
ATSDR believes that the model used for this report is health-protective and will overestimate the actual levels that would be in free-range chicken eggs. One reason why ATSDR believes this is that PCBs were not detected in eggs sampled on Saipan. EPA's Region 9, based in San Francisco, California, issued a report in 2001 on sampling conducted in the Commonwealth of the Northern Mariana Islands in Tanapag Village on the Island of Saipan. This document provided sampling information for a number of different food items (EPA 2001).
- PCBs were not detected in three eggs of free-ranging chickens found near a cemetery, where transformers containing PCB oil were stored, in Tanapag Village on the Island of Saipan. PCBs were not detected in these eggs even though higher soil levels were found in the cemetery over a considerably larger area than were found in the APP West Drainage Area and outfall and residential soils and sediments between APP and the Church.
- The same model used in this document predicted levels of PCBs in eggs from free-ranging chickens in Tanapag Village to be above those predicted for Mongmong. However, PCBs in free-ranging chicken eggs in Tanapag Village were not detected.
To simplify the analysis, ATSDR used either the average or maximum PCB concentration detected in surface soil to estimate the concentrations in the free-range animals. The maximum estimated PCB concentration in chickens (5.5 ppm), eggs (2,197 ppm), and swine (141 ppm) at Agana Springs (Upland) are upper-bound values using very conservative soil exposure assumptions. Although these values are worst-case estimates, it is important to note that any free-range animals will be foraging over large areas that are either not contaminated or contain PCB concentrations in soil that are much lower than the maximum concentration detected. This large foraging area will likely result in much lower PCB concentrations in the meat of free-range animals than what we estimated by using the maximum surface soil concentration..
Human Exposure and Potential Health Effects From Consuming Free-Range Animals or Animal Products in the Mongmong Area Near APP
The most common sources of human exposure to PCBs include fish, meat (e.g., chicken, pork, beef), dairy products, and eggs. The U.S. Food and Drug Administration (FDA) has established an allowable tolerance level of 3 ppm on a fat basis for PCBs in the meat of chickens and in eggs. FDA has not established tolerance values for PCBs in pork or other products from swine.
To provide additional perspective, Table 5 presents concentrations of PCBs that have been detected in some common food items from total diet studies conducted in the United States and New Zealand. Although no actual samples of chickens, eggs, or pork have been collected in the Mongmong area, ATSDR estimated the concentrations in free-range chickens and swine foraging around residential properties near APP (see Appendix D for assumptions and methods used for this analysis). ATSDR selected the residential sampling locations for comparison to total diet studies instead of one of the sampling locations that exceeded ATSDR's screening levels because the residential locations best represented PCB concentrations in surface soil in the Mongmong area. Based on discussions with government of Guam representatives, these residential locations are where people are most likely obtaining most of the free-ranging chickens and swine.
Table 5 shows that the average estimated PCB concentrations in free-range chickens and swine in residential areas adjacent to APP are approximately 1 to 2 orders of magnitude lower than what has been detected in the United States and New Zealand total diet studies. The estimated average PCB concentrations in eggs from Mongmong are about the same as what has been detected in the United States and the estimated maximum PCB concentration was similar to eggs from New Zealand.
ATSDR uses risk model calculations as part of its screening process, which is similar to the risk model calculations EPA uses to support regulatory decisions. Using the estimated concentrations for free-range chickens, chicken eggs, and swine, ATSDR calculated non-cancer and cancer exposure doses from consuming these animals from residential areas, where soil levels were below ATSDR's screening values. In addition, ATSDR calculated non-cancer and cancer exposure doses from consuming these animals from Agana Springs (Upland), APP Drainage Outfall (Upland), and APP West Drainage areas, where past soil levels were above ATSDR's screening values.
ATSDR's Minimal Risk Levels (MRLs)
The MRL is derived from the lowest observed adverse effect level (LOAEL) identified in the scientific literature. The LOAEL is the lowest dose at which an adverse health effect has been observed. It is divided by a safety factor to protect sensitive groups and to account for the differences between humans and animals in response to exposure. MRLs are protective by design and represent doses below which non-cancer adverse health effects are not expected to occur, even from daily exposure over a lifetime. MRLs are not thresholds for harmful health effects. A dose that exceeds the MRL indicates only the increasing potential for toxicity and that further toxicological evaluation is needed.
MRL for PCBs
ATSDR and EPA have evaluated the non-cancer oral toxicity data for PCBs. ATSDR derived a chronic oral MRL of 0.00002 mg/kg/day for PCBs based on data for Aroclor 1254. EPA has not developed a reference dose for PCBs as a class, but has conducted non-cancer assessments for Aroclor 1016, 1048, and 1254.
ATSDR's toxicological evaluation, which includes reviewing animal and human studies and applying realistic assumptions about exposure, indicated that there should be no adverse health effects, either cancer or non cancer. See Appendix D for more information on the exposure dose calculation methods. Most of the estimated non-cancer doses from the residential areas adjacent to APP were at or below ATSDR's health protective minimal risk level (MRL) for PCBs of 0.00002 mg/kg/day (see Appendix D, Table D-3). The highest estimated dose (0.00009 mg/kg/day) for a child consuming free-range eggs in the residential areas is in the same order of magnitude (i.e., 1 in 100,000) as the MRL. Refer to Appendix D for a discussion of the three non-residential off-site sampling locations near APP where PCBs were detected above ATSDR's screening levels.
ATSDR believes that any health concerns associated with consuming foods that contain very low levels of contaminants should be balanced with the benefit of eating a healthy nutritious diet. The U.S. Department of Agriculture's (USDA's) Food Guide Pyramid is a valuable guide that helps provide choices for eating a healthful diet3. According to the Pyramid, an adult should consume two or three servings (one serving = 2 to 3 ounces) of meat, poultry, or fish per day (USDA 2003). These guidelines are consistent with the serving size ATSDR used to estimate the PCB dose from consuming free-range chicken, chicken eggs, and meats from swine in the Mongmong area. ATSDR's assumptions about the average portion of free-range chicken or meat from swine are consistent with USDA's guidance. For example, a person consuming the USDA-recommended daily portion of meat, poultry, and fish would be consuming between 3 and 9 ounces per day. ATSDR used the upper end of this serving size range (i.e., 8.9 ounces) to represent a typical meal size for meat from free-range chickens and pigs. Following this guideline, even if the source of most of their meat and poultry is free-range animals from the residential areas near APP, ATSDR believes that people would not be exposed to PCBs above levels that would be found in a typical U.S. mainland or international diet.
3ATSDR also considered guidelines for traditional Asian diets and evaluated whether estimated doses associated with consuming free-range animal products would pose health concerns based on an Asian Food Pyramid. The Asian Food Pyramid is largely plant-based and recommends small daily servings of low-fat dairy products, only weekly servings of eggs and poultry, and no more than monthly servings of red meat. Since people who follow Asian Food Pyramid guidelines would consume fewer servings and smaller portions of poultry, meats, and eggs, the PCB doses from consuming free-range animal products would be considerably lower than estimated by ATSDR (Cornell University 1995).