Appendix G

Response to Public Comments

The Agency for Toxic Substances and Disease Registry (ATSDR) received the following comments during the public comment period (November 14, 2002 to January 27, 2003) for the Fish and Shellfish Evaluation for the Isla de Vieques Bombing Range Public Health Assessment (PHA) (November 2002). For comments that questioned the validity of statements made in the PHA, ATSDR verified or corrected the statements. The list of comments does not include editorial comments, such as word spelling or sentence syntax.

1. Comment: Between this fish and shellfish evaluation and the earlier groundwater and drinking water pathways et al studies I consider the inhabitants of Isla de Vieques to be in a reasonably safe environment.

   Response: Thank you for the comment. After thoroughly evaluating the drinking water and groundwater supplies, the soil pathway, the fish and shellfish, and the air pathway, ATSDR agrees that the residents of Vieques are not being exposed to harmful levels of chemicals from the Live Impact Area (LIA).

2. Comment: While stated in different parts of the document, the document summary should name the metals found in the fish and shellfish and conclude that while the following compounds were found to be present in the fish and shellfish of Vieques, they are found at levels low enough as not to pose a threat to human health. The individual metals found should also be included in Part VIII, Conclusion.

   Response: ATSDR agrees with the statement that while metals were detected in the Vieques fish and shellfish, they were found at levels too low to present a public health concern. However, since very few metals were not detected in Vieques fish and shellfish, listing all of them would not be efficient. Instead, please refer to Tables 8 and 9 which identify the ranges, averages, and frequency of detections for the metals that were detected in the fish and shellfish, respectively. It should be noted that heavy metals are commonly detected in fish and shellfish tissue because seafood tends to accumulate metals that are naturally present in the environment.

3. Comment: The report should have one or two sentences stating that the findings of this report are directed solely at human health and cannot be used to infer ecological integrity of the natural systems on Vieques. The results of an ATSDR report should not be used to imply that there are no environmental impacts on Vieques.

   Response: ATSDR agrees that this evaluation is a public health assessment, not an ecological assessment.

4. Comment: The importance of how explicit appear the text is to determine the existence or not of the contamination and the pathways.

   Response: ATSDR agrees that it is important to explicitly present the findings about the existence of contamination and the possibility of exposure pathways.

5. Comment: The PHA is strongly based on Dr. Caro's research, Metropolitan University. While Dr. Caro's report determines that her research is a starting point; inclusive without having access to the complete methodology of the research, I question that epidemiologically, 51 residents are a representative percentage of the total population of Vieques Island.

   Response: At the time of ATSDR's sampling, Dr. Caro's research provided the most detailed and relevant information about consumption habits of Vieques residents (Caro et al. 2000). However, it was not the only source consulted by ATSDR. We also relied on information provided by (1) the petitioner, (2) the Vieques Special Commission Report (Government of Puerto Rico 1999 as cited in Navy 2000b), (3) ATSDR's community involvement program, and (4) visits to local fish markets by ATSDR staff. Taken together, these sources provided ATSDR with a good indication of how often residents of Vieques eat fish and shellfish and what species they eat.

6. Comment: Epidemiologically does not establish the absence of risk.

   Response: ATSDR assumes that this comment is in reference to Dr. Caro's research of a representative sampling (51 people) of Vieques residents' consumption habits (Caro et al. 2000). Dr. Caro's research helped ATSDR determine which species of fish and shellfish to collect and helped determine how often residents of Vieques ate fish and shellfish. The information provided in the report was substantiated by information collected by the petitioner, ATSDR's community involvement program, the Vieques Special Commission Report (Government of Puerto Rico 1999 as cited in Navy 2000b), and visits to local fish markets.

   To determine whether it is safe to eat the fish and shellfish from around Vieques, ATSDR collected 104 fish and 38 shellfish and analyzed them for heavy metals and explosive compounds. The results of the sampling and analysis indicated what levels of chemicals are present in the fish and shellfish around Vieques. ATSDR used these data to determine that it is safe for people to eat fish and shellfish with the detected levels of chemicals.

7. Comment: The methodology by which the nutritional information is obtained is unclear.

   Response: To determine which species of commonly consumed fish should be sampled, ATSDR compiled consumption information from the questionnaire administered by the Universidad Metropolitana researchers (Caro et al. 2000), the petitioner, the Vieques Special Commission Report (Government of Puerto Rico 1999 as cited in Navy 2000b), ATSDR's community involvement program, and visits to local fish markets by ATSDR staff.

   1. Some of the information regarding the relative frequency of consumption by Viequenses was obtained from the study done by Metropolitan University (Caro et al. 2000). As quoted by the ATSDR, a total of 51 Viequenses was surveyed in Dr. Caro's study and the frequency of fish consumption was derived from this survey. The age and sex distribution of the 51 subjects is not known.

   Response: The 51 residents of Vieques who were questioned by Universidad Metropolitana were comprised of 20 females (39.2%) and 31 males (60.8%). While the researchers did not report the actual ages of the residents who participated in the questionnaire, they did ask how long they lived in Vieques. Eleven people (21.5%) reported that they lived in Vieques 1-10 years, 9 people (17.6%) lived in Vieques 11-20 years, 10 people (19.6%) lived in Vieques 21-35 years, and 21 people (41.2%) lived in Vieques 36 or more years. This information is provided in the original report (Caro et al. 2000).

   1. The potential consumption of seafood (fish, mollusks, and/or crustaceans) more than once daily is not contemplated in the Dr. Caro study:

   fish consumed % (of subjects)	frequency
   47.05	1-2 times per week
   17.60	3-4 times per week
   15.60	5 or more times per week
   19.60	0 or occasional

   Response: It appears that the Universidad Metropolitana researchers asked how many times a week the residents ate seafood. This seems to imply that a person could respond with a number greater than seven if they ate fish or shellfish more than once a day. The category of "5 or more times per week" would; therefore, include anyone who eats fish or shellfish more than once a day.

8. Comment: The methodology and questionnaire by which ATSDR interviewed Viequenses regarding their nutritional habits are not stated.

   Response: ATSDR did not formally interview or circulate questionnaires to the residents of Vieques asking about their nutritional habits. To determine which species of commonly consumed fish should be sampled, ATSDR compiled consumption information from the questionnaire administered by the Universidad Metropolitana researchers (Caro et al. 2000), the petitioner, the Vieques Special Commission Report (Government of Puerto Rico 1999 as cited in Navy 2000b), ATSDR's community involvement program, and visits to local fish markets by ATSDR staff.

9. Comment: Our study done during 2000-2002, included among other subjects, a detailed questionnaire of nutritional habits. These were completed by 102 volunteers. Among our findings are the following:

   1. Seafood (fish, mollusks and crustaceans) consumption is highly variable.

   2. The frequency of fish consumption varies significantly with the age variable.

   3. The weekly total seafood consumption for the 102 studied Viequenses was 5.9 +/- 1.1 times per week. The upper limit of the 95% confidence interval was 9.7 times per week. Thus some Viequenses consume seafood more frequently than once daily.

   4. In contrast, weekly chicken consumption average 3.5 +/- 0.2 times per week, with a 95% CI limit of 3.8 and did not exhibit a statistically significant correlation with the age variable.

   Response: ATSDR requested, but was not given, the opportunity to view this particular study. ATSDR and US Environmental Protection Agency's Environmental Response Team (EPA/ERT) collected commonly caught and consumed species based on several sources of information that were available at the time of the sampling (prior to July 2001): (1) the questionnaire administered by the Universidad Metropolitana researchers (Caro et al. 2000), (2) the petitioner, (3) the Vieques Special Commission Report (Government of Puerto Rico 1999 as cited in Navy 2000b), (4) ATSDR's community involvement program, and (5) visits to local fish markets by ATSDR staff. Taken together, these sources provided ATSDR with a good indication of how often residents of Vieques eat fish and shellfish and what species they eat.

   According to the consumption information provided by the commenter, the majority of residents of Vieques were considered during ATSDR's evaluation of eating fish and shellfish once a day, every day for 70 years. To evaluate whether the small percentage of people who eat fish and shellfish 9.7 times a week would expect to experience adverse health effects, ATSDR adjusted the exposure frequency to 505 days a year in the exposure dose equations provided in Appendix D. The results are provided in the following table. All of these doses are below levels of health concern for anyone eating fish and shellfish 9.7 times a week.

   Chemical	Fish			Shellfish
   	Average Concentration (ppm)	Estimated Exposure Dose (mg/kg/day)		Average Concentration (ppm)	Estimated Exposure Dose (mg/kg/day)
   		Adult	Child		Adult	Child
   Inorganic Arsenic	4.0	0.0036	0.0079	11.1	0.0100	0.0218
   Cadmium	0.08	0.0004	0.0008	0.36	0.0016	0.0035
   Chromium	0.16	0.0007	0.0016	0.19	0.0009	0.0019
   Lead	0.27	0.0012	0.0026	0.25	0.0011	0.0025
   Mercury	0.12	0.0005	0.0012	0.03	0.0001	0.0003
   Selenium	0.98	0.0044	0.0096	0.80	0.0036	0.0079
   Zinc	3.1	0.014	0.031	30.1	0.135	0.295
   Chemical	Average Concentration (ppm)	Estimated Daily Dose (mg/day)		Average Concentration (ppm)	Estimated Daily Dose (mg/day)
   Copper	0.56	0.18	0.09	7.8	2.5	1.2
   Iron	6.8	2.2	1.1	36.3	11.5	5.8
   Chemical	Average Concentration (ppm)	Estimated Blood Lead Level (mg/dl)		Average Concentration (ppm)	Estimated Blood Lead Level (mg/dl)
   Lead	0.27	0.09	0.09	0.25	0.08	0.08

10. Comment: The rationale for sampling the chosen species and the number of individuals sampled for each species has not been determined. To our knowledge, prior to our study, the frequency or relative frequency of local consumption of each species had not been determined. For illustration, a rank of the most commonly consumed species is provided. The number of individuals sampled for each species (as per ATSDR reports) is also listed.

    Species	sampled by ATSDR	sampled by UMET
    colirubia(Ocyurus chrysurus)	10	12
    chapin (Lactophyrys trigonus)	1	0
    sierra (Scomberomorus regalis)	0	0
    carrucho (Strombus gigas)	4	0
    pulpo (Octopus vulgaris)	0	0
    caracol-Burgao(Cittarium pica)	0	0
    sama (lutjamus analis)	0	0
    "chopa" (multiple species)	information not available	information not available
    tiburon (Shark) (several species)	0	0
    pargo (Lutjanus apodus)	0	0
    chillo (lutjanus vivanus)	0	0
    rabalo	N/A	N/A

    Thus, the only commonly consumed species for which more than 3 individuals were sampled was colirubia (Ocyurus chrysurus). Any conclusion about other species or variety of seafood is unfounded.

    Response: ATSDR requested, but was not given the opportunity to view this particular study. ATSDR and EPA/ERT collected commonly caught and consumed species based on several sources of information that were available at the time of the sampling (prior to July 2001): (1) the questionnaire administered by the Universidad Metropolitana researchers (Caro et al. 2000), (2) the petitioner, (3) the Vieques Special Commission Report (Government of Puerto Rico 1999 as cited in Navy 2000b), (4) ATSDR's community involvement program, and (5) visits to local fish markets by ATSDR staff.

    It should be noted that the table provided by the commenter is inaccurate in the number of fish ATSDR collected for each species identified. As Table 7 shows, ATSDR collected more than three individuals for three of the species specifically identified by the commenter (colirubia, carrucho, and pargo). During the July 2001 sampling, ATSDR and EPA/ERT collected 104 fish and 38 shellfish, representing 17 different species of fish and three different species of shellfish--30 groupers (Epinephelus sp.), 19 snappers (Ocyurus chrysurus and Lutjanus sp.), 25 parrotfish (Scaridae family), 24 grunts (Haemulon sp.), 5 goatfish (Mullidae family), 1 cowfish (Lactophrys polygonia) , 20 conch (Strombus gigas), 7 lobster (Panulirus argus), and 11 land crab (Cardisoma guanhumi) composite samples.

    Because not every species of fish and shellfish can be realistically sampled (nor is it expected or recommended⁽¹²⁾), ATSDR and EPA/ERT focused the sampling effort on species of commonly consumed fish and shellfish that would be expected to have a higher exposure to potential contamination from Navy training exercises at the LIA. Therefore, ATSDR and EPA/ERT choose to sample non-migratory, reef-dwelling species that tended to reside locally (rather than migratory, open-water fish that would have less exposure to local contamination).

11. Comment: The distinction between seafood locally consumed and seafood fished for commercial purposes in Vieques is not clarified.

    Response: The purpose of ATSDR's sampling and analysis activities was to determine whether the muscle tissues from commonly consumed fish and shellfish, collected from the coastal waters and near shore land on Vieques, contain levels of heavy metals and explosives compounds that would adversely affect public health. ATSDR's public health evaluation assesses potential exposures from eating fish and shellfish from Vieques, regardless of where a person is when they eat the fish and shellfish.

12. Comment:The primary research question being addressed is whether military operations on the island have contaminated fish consumed by Island residents. The ATSDR sampling design employed to detect chemicals in fish did not produce data that would allow ATSDR to answer this question with statistical significance. The ideal sampling design would grow from an understanding of where chemicals have been released to the environment, their movement, and fate. It would permit the testing of hypotheses regarding relations between fish contamination levels and their proximity to known sites of release and contamination. The design would further test the hypothesis that certain species of fish may accumulate some elements or compounds more efficiently than others. It should also consider the behavior patterns of the various species, and their tendency to remain proximate to contaminated areas, or to migrate longer distances, perhaps to cleaner waters. The ATSDR choice of "productive fishing areas surrounding the island" (p. 18) as a criterion for sampling is necessary but insufficient to understand possible associations between chemical concentrations in fish and military contamination. The ATSDR sampling effort is especially difficult to understand absent a map of the Island and offshore areas, designating sampling locations, numbers of fish and shellfish samples collected at each location. Current maps do not provide this information.

    Response: The purpose of ATSDR's sampling and analysis activities was to determine whether the muscle tissues from commonly consumed fish and shellfish contain levels of heavy metals and explosives compounds that would adversely affect public health. Therefore, ATSDR's sampling and analysis was sufficient for the public health evaluation that was conducted. While it may be a useful environmental/ecological exercise to conduct a sampling program similar to the one described by the commenter, it would have been a strictly academic exercise. The primary purpose of determining that the fish and shellfish are safe to eat was accomplished--the levels of metals that were detected in the fish and shellfish are below levels of health concern.

    During the July 2001 sampling, ATSDR and EPA/ERT collected 104 fish and 38 shellfish, representing 17 different species of fish and three different species of shellfish (see Table 7). The types of fish and shellfish collected were previously identified through several sources as species that are commonly caught and consumed by the residents of Vieques. This methodology is in accordance with EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories (EPA 2000).

    Because not every species of fish and shellfish can be realistically sampled (nor is it expected or recommended⁽¹³⁾), ATSDR and EPA/ERT focused the sampling effort on species of commonly consumed fish and shellfish that would be expected to have a higher exposure to potential contamination from Navy training exercises at the LIA. Therefore, ATSDR and EPA/ERT choose to sample non-migratory, reef-dwelling species that tended to reside locally (rather than migratory, open-water fish that would have less exposure to local contamination).

    ATSDR agrees that, for a variety of reasons, different areas around Vieques may contain varying levels of chemicals. To evaluate whether eating fish and shellfish from different locations around Vieques would adversely affect public health, ATSDR and EPA/ERT caught fish and shellfish from five "preferred" fishing locations around Vieques--north of the LIA, south of the LIA, south of Esperanza, north of Isabel Segunda, and west of the former NASD--as well as from a local fish market in Isabel Segunda (see Figure 3). Table 7 provides the types and number of species collected at each location.

    The fish and shellfish collected from around the LIA (areas of presumed highest contamination) do not contain levels of metals or explosives compounds that would adversely affect the health of someone eating fish and shellfish from this area (see the Public Health Evaluation (Section VB) and Appendix D for more details). In addition, despite the common occurrence of unexploded ordnance, the site was home to a diverse population of apparently healthy fish and small head corals, and were surrounded by a large halo (the halo is a typical feature of reefs and underwater structures that is caused by grazing organisms leaving the shelter of the reef at night to feed on the surrounding seagrass beds) and healthy turtle grass bed (see Appendix C for more details).

13. Comment: Contaminants in fish are likely to be higher than recorded by ATSDR's sampling and analyses, if fish are collected closer to the time, and nearer the location of military activity. Sampling efforts should be timed to follow intense periods of bombardment, when hazardous chemicals are released and mobilized. Rains, winds, currents and tides will all work to dilute pollutants across space and time.

    Response: Because ATSDR's involvement began in May 1999, after the Navy ceased military training with live munition, it was not possible for fish and shellfish to be collected during live bombing exercises. From July 16-20, 2001, ATSDR and EPA/ERT collected commonly consumed fish and shellfish from six locations on Vieques. The Navy conducted training exercises with inert ordnance from June 12-29, 2001, less than three weeks before ATSDR and EPA/ERT's sampling (personal communication with Navy personnel, February 2003).

    ATSDR agrees that dilution will decrease the likelihood of exposure at some point in time and distance from the bombing range. Therefore, ATSDR sampled two locations immediately to the north and south of the LIA, areas of presumed highest contamination. Fish and shellfish from these two locations did not show levels high enough to be of health concern.

14. Comment: The ATSDR sample sizes, admittedly constrained by available time and financial resources, are nevertheless grossly insufficient to capture likely variance in chemical concentrations across space, time, and among species. For example only 11 yellowtail snappers were collected by ATSDR, despite the fact that it was estimated to be the most commonly consumed fish by island residents (Caro et al. 2000). The most commonly consumed shellfish include blue land crab (n=11), spiny lobster (n=7) and queen conch (n=20) where n is the number of individual fish tested. Since elements and compounds are likely to behave differently in different species, the small ATSDR sample sizes are clearly insufficient to predict exposures, and to reach conclusions regarding the safety or absence of health hazard associated with chronic human fish consumption.

    Response: During the July 2001 sampling, ATSDR and EPA/ERT collected 104 fish and 38 shellfish, representing 17 different species of fish and three different species of shellfish--30 groupers, 19 snappers, 25 parrotfish, 24 grunts, 5 goatfish, 1 cowfish, 20 conch, 7 lobster, and 11 land crab composite samples (see Table 7). The types of fish and shellfish collected were previously identified through several sources as species that are commonly caught and consumed by the residents of Vieques. This methodology is in accordance with EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories (EPA 2000).

    Because the purpose of ATSDR's sampling and analysis activities was to determine whether commonly consumed fish and shellfish contain levels of heavy metals and explosives compounds that would adversely affect public health, ATSDR chose to sample a "typical sized species" as identified at the market and in the waters on the reefs near the island. Since Navy bombing activities are conducted on the LIA, ATSDR sampled two locations immediately to the north and south of the LIA, areas of presumed highest contamination.

15. Comment: The Agency claims that its calculation of averages using only detects results in a conservative overestimation of health effects. This may not be the case when the data are disaggregated by chemical and fish species, which will reduce the sample size. It is also likely that if the Agency sampled more intensively, considering factors such as time-from-release date, proximity to release, and species behavior, the high end of detects would be substantially higher than those reported. By not releasing individual data points the public is left uncertain regarding the shape of the distribution of findings, and this knowledge is critical to estimate probable human exposure levels.

    Response: There are two general methods to incorporate nondetected values--one way is to include them as zero and the other is to include them using one half the detection limit. Either way, the average would be lower than the number used by ATSDR to calculate exposure doses since ATSDR only used values higher than the detection limits (i.e., detected values). Since the detected concentrations within a group fell within the expected or normal distribution, it is likely that the average sample of fish collected is representative of the rest of the population (of that fish size).

    The Navy conducted training exercises with inert ordnance from June 12-29, 2001, less than three weeks before ATSDR and EPA/ERT's fish and shellfish sampling activities from July 16-20, 2001 (personal communication with Navy personnel, February 2003).

    ATSDR and EPA/ERT focused the sampling effort on species of commonly consumed fish and shellfish that would be expected to have a higher exposure to potential contamination from Navy training exercises at the LIA by choosing non-migratory, reef-dwelling species that tended to reside locally (rather than migratory, open-water fish that would have less exposure to local contamination). In addition, ATSDR sampled two locations immediately to the north and south of the LIA, areas of presumed highest contamination.

16. Comment: Actual levels of methylmercury in Vieques fish may be higher than reported levels due to the type of fish sampled and the small sample size surveyed by ATSDR.

    Response: ATSDR and EPA/ERT sampled commonly consumed fish and shellfish that would be expected to have a higher exposure to potential contamination from Navy training exercises at the LIA by choosing non-migratory, reef-dwelling species that tended to reside locally (rather than migratory, open-water fish that would have less exposure to local contamination).

    In general, the levels of mercury measured in fish collected from Vieques were about the same as those from the mainland of Puerto Rico and the Virgin Islands. In Vieques, the average mercury level was 0.12 ppm. Average mercury levels found in the same species of fish from Puerto Rico and the Virgin Islands ranged from 0.07 to 0.70 (Burger et al. 1992). Average mercury levels in seafood species, collected nationwide, ranged from not detected to 1.45 ppm (FDA 2001b). It is also interesting to note that people who eat fish from Vieques would receive about as much mercury as people who eat canned tuna (according to a 1991 FDA survey, the average mercury concentration in canned tuna is 0.17 ppm; Yess 1993 as cited in ATSDR 1999a).

    ATSDR and EPA/ERT sampled 104 fish and 38 shellfish. Collecting more fish will not likely increase the levels of methylmercury found in the Vieques fish and shellfish. While it is probable that some fish may have higher levels, it is also equally likely that some fish would contain lower levels. Since the detected concentrations within a group fell within the expected or normal distribution, it is likely that the average sample of fish collected is representative of the rest of the population (of that fish size).

17. Comment: Despite limitations in sampling design and sample size, ATSDR reached three aggressive and I believe unsupportable conclusions:

    • "It is safe to eat a variety of fish and shellfish every day."
    • "It is safe to eat fish and shellfish from any of the locations sampled, including from around the LIA and the sunken Navy target vessel."
    • "It is safe to eat the most commonly consumed species, snapper, every day." (ATSDR 2003 pp. 2-3).

    It is my opinion that the data presented do not justify ATSDR's conclusions that fish intake by Vieques residents poses no health threat. There is evidence of significant contamination of the Vieques landscape and marine ecosystems, resulting from intensive U.S. military activities over the past 60 years. Does ATSDR's surveillance demonstrate safety? Unequivocally, the answer is no. The absence of evidence should not be used by ATSDR to claim the absence of a health hazard, given important defects in the Agency's environmental surveillance. A far more ambitious sampling effort, stratified by species, location and time would be necessary before conclusions such as those presented in this report could be viewed as credible science.

    Response: ATSDR disagrees. The purpose of ATSDR's sampling and analysis activities was to determine whether the muscle tissues from commonly consumed fish and shellfish contain levels of heavy metals and explosives compounds that would adversely affect public health. ATSDR's sampling and analysis was sufficient for determining that the fish and shellfish are safe to eat.

    Heavy metals were found in the tissues of the fish and shellfish collected by ATSDR and EPA/ERT (see Tables 8 and 9). This is not uncommon--heavy metals are commonly detected in fish and shellfish tissue because seafood tends to accumulate metals that are naturally present in the environment. However, all of the concentrations were detected at levels too low to present a public health concern to people who eat fish and shellfish from Vieques.

    ATSDR and EPA/ERT collected 104 fish and 38 shellfish, representing 17 different species of fish and three different species of shellfish--30 groupers, 19 snappers, 25 parrotfish, 24 grunts, 5 goatfish, 1 cowfish, 20 conch, 7 lobster, and 11 land crab composite samples (see Table 7). The types of fish and shellfish collected were previously identified through several sources as species that are commonly caught and consumed by the residents of Vieques. To evaluate whether eating fish and shellfish from different locations around Vieques would adversely affect public health, ATSDR and EPA/ERT caught fish and shellfish from five "preferred" fishing locations around Vieques--north of the LIA, south of the LIA, south of Esperanza, north of Isabel Segunda, and west of the former NASD--as well as from a local fish market in Isabel Segunda (see Figure 3). This methodology is in accordance with EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories (EPA 2000).

    The Navy conducted training exercises with inert ordnance from June 12-29, 2001, less than three weeks before ATSDR and EPA/ERT's fish and shellfish sampling from July 16-20, 2001 (personal communication with Navy personnel, February 2003).

18. Comment: Why were no analyses conducted to search for the presence of radionuclides?

    Response: Radiological contamination around the former USS Killen was not an issue when ATSDR initiated fish and shellfish sampling activities at Vieques. After the concern was raised, ATSDR spoke with a Navy radiation scientist about the presence of radioactivity and evaluated radiation measurements from 1975 and 2002. The radiation levels associated with the former USS Killen are indistinguishable from the radiation associated with background and do not pose any public health hazard to the residents of Vieques. Please refer to the Community Health Concerns section (Section VI) for a discussion of the concern about possible radiological contamination around the former USS Killen.

19. Comment: Table D-3 (Estimated exposure doses from ingestion of fish), pages D-36 to 37 and Table D-4 (Estimated exposure doses from ingestion of shellfish) pages D-38 to 40, shows contaminants that exceed the health guides; even with that ATSDR justify itself saying that this does not necessarily result in an adverse health effect.

    Response: Table D-3 and Table D-4 within the PHA compare estimated exposure doses to the oral health guidelines (e.g., minimal risk levels (MRLs) and reference doses (RfDs)). In these tables ATSDR identified which chemicals required further evaluation, noted with an asterisk. However, as mentioned at the end of the tables, "An exposure dose that is higher than the MRL or RfD does not necessarily result in harmful health effects. These metals are further evaluated." Exposure to a level above the MRL or RfD does not mean that adverse health effects will occur. MRLs and RfDs are intended only to serve as screening tools to help public health professionals decide whether to conduct a more comprehensive assessment of potential health effects. This concept is further explained in the Using exposure doses to evaluate potential health hazards section in Appendix D.

20. Comment: The RfD is defined as the estimated daily consumption that is likely to be without deleterious effects during the lifetime for humans (2). MRL is not pertinent to mercury as this is not a known carcinogen.

    Response: Both the MRL and RfD are estimates of daily human exposure to a substance that are unlikely to result in noncancer effects over a specified duration. Therefore, the MRL is applicable to mercury as a noncarcinogen. Definitions of the MRL and RfD can be found in Appendix A, ATSDR Glossary of Environmental Health Terms.

21. Comment: The U.S. National Academy of Sciences (NAS) concluded in 2000 that the most scientifically defensible RfD for human consumption of methylmercury is 0.1 ug/kg/day. The ATSDR MRL is 0.3 ug/kg/day, permitting a level 3 times higher than the NAS supports. ATSDR should explain the knowledge they possess that justifies their conclusion that humans may be safely exposed to methylmercury at levels 3 times higher than the NAS recommends.

    Response: ATSDR's MRLs undergo a rigorous review process: Health Effects/MRL Workgroup reviews within ATSDR's Division of Toxicology, expert panel peer reviews, and agency wide MRL Workgroup reviews, with participation from other federal agencies and comments from the public. MRLs are derived for hazardous substances using the no-observed-adverse-effect level (NOAEL)/uncertainty factor approach. The derivation for mercury is explained in detail within Appendix A of ATSDR's Toxicological Profile for Mercury (ATSDR 1999a).

    The chronic oral MRL for methylmercury is based on the Seychelles Child Development Study, in which over 700 mother-infant pairs have, to date, been followed and tested from parturition through 66 months of age. The Seychellois regularly consume a large quantity and variety of ocean fish, with 12 fish meals per week representing a typical methylmercury exposure. In this study, developing fetuses were exposed to methylmercury in utero through maternal fish ingestion before and during pregnancy. Neonates continued to be exposed to maternal mercury during breast feeding, and methylmercury exposure from the regular diet continued after the gradual post-weaning shift to a fish diet. The results revealed no evidence of adverse effects attributable to chronic ingestion of low levels of methylmercury in fish (Davidson et al. 1998 as cited in ATSDR 1999a). ATSDR derived the chronic-duration MRL for mercury (0.0003 mg/kg/day) by dividing the NOAEL from the Seychelles Child Development Study (0.0013 mg/kg/day) by an uncertainty factor of 3 to account for human pharmacokinetic and pharmacodynamic variability and by a modifying factor of 1.5.

22. Comment: Table D-3 reveals that the RfD for mercury (0.0001 mg/kg/day) was exceeded even if an individual (child - 0.00033 or adult - 0.00015) consumed fish only once daily according to ATSDR samples. This is 330% and 150% higher than the sanitary standard.

    Response: Table D-3 compares the estimated exposure doses from ingestion of fish collected by ATSDR to the chemical-specific oral health guidelines. In this table ATSDR identified which chemicals required further evaluation, noted with an asterisk. However, as mentioned at the end of the table, "An exposure dose that is higher than the MRL or RfD does not necessarily result in harmful health effects. These metals are further evaluated." Exposure to a level above the MRL or RfD does not mean that adverse health effects will occur. MRLs and RfDs are intended only to serve as screening tools to help public health professionals decide whether to conduct a more comprehensive assessment of potential health effects. This concept is further explained in the Using exposure doses to evaluate potential health hazards section in Appendix D. ATSDR further evaluated mercury exposure (as well as the other chemicals identified with an asterisk) in both children and adults within the Mercury section of Appendix D. Table D-14 presents the exposure doses expected to result from eating fish and shellfish from Vieques on a daily basis. All of the estimated exposure doses were below toxicological doses reported in the scientific literature in which no adverse health effects were reported. Therefore, the levels of mercury found in the fish and shellfish collected from Vieques are not of health concern.

    As a note, all of the fish and shellfish collected from Vieques contained levels of total mercury below the US Food and Drug Administration's (FDA's) guidance for methylmercury in fish, shellfish, and crustacea (1 ppm). The maximum concentration of total mercury in fish was a third of this guidance (maximum = 0.33 ppm, see Table 8), while the maximum concentration in shellfish was even lower (maximum = 0.049 ppm, see Table 9).

23. Comment: Table D-20 revealed that the maximum detected mercury fish concentrations resulted in exposure dosages exceeding the RfD for mercury (child 0.0027 and adult 0.0012). Similarly, the child exposure exceeds the RfD by 27000% and the adult exposure will exceed the RFD by 1200%.

    Response: Table D-20 compares the estimated exposure doses from ingestion of fish collected by Universidad Metropolitana to the chemical-specific oral health guidelines. In this table ATSDR identified which chemicals required further evaluation, noted with an asterisk. However, as mentioned at the end of the table, "An exposure dose that is higher than the MRL or RfD does not necessarily result in harmful health effects. These metals are further evaluated." Exposure to a level above the MRL or RfD does not mean that adverse health effects will occur. MRLs and RfDs are intended only to serve as screening tools to help public health professionals decide whether to conduct a more comprehensive assessment of potential health effects. This concept is further explained in the Using exposure doses to evaluate potential health hazards section in Appendix D. ATSDR further evaluated adult and child exposures to all of the chemicals with an asterisk, including mercury, within Appendix D, specifically within the Evaluating Health Concerns Using Fish Fillet Data from Universidad Metropolitana section. Table D-21 presents the exposure doses expected to result from eating fish on a daily basis. All of the estimated exposure doses were below toxicological doses reported in the scientific literature in which no adverse health effects were reported. Therefore, the levels of mercury found in the fish collected by Universidad Metropolitana from the fish markets in northern Vieques, Esperanza, and the Parquera area on the mainland of Puerto Rico are not of health concern.

    As a note, all of the fish collected by Universidad Metropolitana contained levels of mercury below FDA's guidance for methylmercury in fish, shellfish, and crustacea (1 ppm). The maximum concentration was about a third of this guidance (maximum = 0.38 ppm, see Tables 3 and 4).

24. Comment: The RfD (and MRL) for arsenic oral consumption is 0.0003 mg/kg/day. Table D-3 revealed that ATSDR sampled fish will produce exposure that by far exceed the RfD.

    Child 0.03/0.0003 = 100 times higher or 10000%
    
    Adult 0.014/0.0003 = 46.7 times higher or 4,667%

    Response: Table D-3 compares the estimated exposure doses from ingestion of fish collected by ATSDR to the chemical-specific oral health guidelines. In this table ATSDR identified which chemicals required further evaluation, noted with an asterisk. However, as mentioned at the end of the table, "An exposure dose that is higher than the MRL or RfD does not necessarily result in harmful health effects. These metals are further evaluated." Exposure to a level above the MRL or RfD does not mean that adverse health effects will occur. MRLs and RfDs are intended only to serve as screening tools to help public health professionals decide whether to conduct a more comprehensive assessment of potential health effects. This concept is further explained in the Using exposure doses to evaluate potential health hazards section in Appendix D. ATSDR further evaluated arsenic exposure (as well as the other chemicals identified with an asterisk) in both children and adults within the Arsenic section of Appendix D. Tables D-5 and D-6 present the exposure doses expected to result from eating fish and shellfish from Vieques on a daily basis. All of the estimated exposure doses were within the body's capability to metabolize arsenic. Therefore, the levels of arsenic found in the fish and shellfish collected from Vieques are not of health concern.

    As a note, all of the fish and shellfish collected from Vieques contained levels of arsenic below FDA's guidance for arsenic in crustacea (76 ppm) and molluscan bivalves (86 ppm). The maximum concentration in fish was about a third of this guidance (maximum = 29.3 ppm, see Table 8) and the maximum concentration in shellfish was also below this guidance (maximum = 48.3 ppm, see Table 9).

25. Comment: Table D-20 revealed that ATSDR sampled fish will produce exposure that by far exceed the RfD for arsenic.

    Child 0.0102/0.0003 = 34 times higher or 3,400%
    
    Adult 0.0047/0.0003 = 15.7 times higher or 1,570%

    Response: Table D-20 compares the estimated exposure doses from ingestion of fish collected by Universidad Metropolitana to the chemical-specific oral health guidelines. In this table ATSDR identified which chemicals required further evaluation, noted with an asterisk. However, as mentioned at the end of the table, "An exposure dose that is higher than the MRL or RfD does not necessarily result in harmful health effects. These metals are further evaluated." Exposure to a level above the MRL or RfD does not mean that adverse health effects will occur. MRLs and RfDs are intended only to serve as screening tools to help public health professionals decide whether to conduct a more comprehensive assessment of potential health effects. This concept is further explained in the Using exposure doses to evaluate potential health hazards section in Appendix D. ATSDR further evaluated adult and child exposure to all of the chemicals with an asterisk, including arsenic, within Appendix D, specifically within the Evaluating Health Concerns Using Fish Fillet Data from Universidad Metropolitana section. Table D-21 presents the exposure doses expected to result from eating fish on a daily basis. All of the estimated exposure doses were within the body's capability to metabolize arsenic and are below the noncancer and cancer health effects levels reported in the scientific literature. Therefore, the levels of arsenic found in the fish collected by Universidad Metropolitana from the fish markets in northern Vieques, Esperanza, and the Parquera area on the mainland of Puerto Rico are not of health concern.

    As a note, all of the fish collected by Universidad Metropolitana contained levels of arsenic below FDA's guidance for arsenic in crustacea (76 ppm) and molluscan bivalves (86 ppm). The maximum concentration was only a fraction of this guidance (maximum = 10 ppm, see Tables 3 and 4).

26. Comment: Sanitary standards are established as guidelines. In the case of mercury, this guideline is based on the concentrations that are neurotoxic to the developing human fetus. Exposures at or below the standard are believed to be without deleterious health effects. The safety of exposures ABOVE the standard has not been determined. Conclusions about the safety of exposures above the RfD are unfounded. Given the magnitude of the potential exposures even at low frequency of consumption by Viequenses, the conclusions of the ATSDR report are unfounded and contradictory.

    Response: MRLs and RfDs are intended only to serve as screening tools to help public health professionals decide whether to conduct a more comprehensive assessment of potential health effects. To maximize human health protection, MRLs and RfDs have built in uncertainty or safety factors, making these values considerably lower than levels at which health effects have been observed. The result is that even if an exposure dose is higher than the MRL or RfD, it does not necessarily follow that harmful health effects will occur. This concept is further explained in the Using exposure doses to evaluate potential health hazards section in Appendix D.

    If health guideline values (MRLs and RfDs) are exceeded, ATSDR examines the health effects levels discussed in the scientific literature and more fully reviews exposure potential. This information is used to describe the disease-causing potential of a particular chemical and to compare site-specific dose estimates with doses shown in applicable studies to result in illness (known as the margin of exposure). This process enables ATSDR to weigh the available evidence in light of uncertainties and offer perspective on the plausibility of harmful health outcomes under site-specific conditions.

    The chronic oral MRL for methylmercury is based on the Seychelles Child Development Study, in which over 700 mother-infant pairs have, to date, been followed and tested from parturition through 66 months of age. The Seychellois regularly consume a large quantity and variety of ocean fish, with 12 fish meals per week representing a typical methylmercury exposure. The results revealed no evidence of adverse effects attributable to chronic ingestion of low levels⁽¹⁴⁾ of methylmercury in fish (NOAEL = 0.0013 mg/kg/day). In this study, developing fetuses were exposed in utero through maternal fish ingestion before and during pregnancy. Neonates continued to be exposed to maternal mercury during breast feeding, and methylmercury exposure from the regular diet continued after the gradual post-weaning shift to a fish diet (Davidson et al. 1998 as cited in ATSDR 1999a). All of the exposure doses estimated by ATSDR for daily consumption of the average concentrations of mercury in fish and shellfish from Vieques were lower than the level at which no adverse health effects were observed in this long-term, subsistence-based study (i.e., lower than the NOAEL of 0.0013 mg/kg/day) (see Table D-14). Therefore, it can confidently be concluded that the mercury concentrations found in the fish and shellfish from Vieques are not at levels of health concern.

27. Comment: In the case of mercury, the ATSDR at no point refers to the criteria published since 1999 by: National Research Council, US Environmental Protection Agency (EPA) Centers for Disease Control and Prevention (CDC), Surgeon General of the United States, and FDA. Namely, recommending women of reproductive age in the USA to not consume more than 6 ounces of fish per week to decrease exposure to mercury to the developing fetus. This is an overt contradiction to ATSDR conclusions (page 16 English/18 Spanish versions):

    "Based on these data, ATSDR determined that it is safe to eat fish from the fish market in northern Vieques, Esperanza, and the Parquera on the mainland of Puerto Rico on a daily basis."

    Response: FDA states that pregnant women and women of child-bearing age can safely eat 12 ounces per week of cooked fish, however, "some kinds of fish are known to have much lower than average levels of methylmercury and can be safely eaten more frequently and in larger amounts." FDA recommends contacting an appropriate food safety authority for specific consumption recommendations about fish caught or sold in a person's local area (FDA 2001a). CDC and EPA concur with this guidance (EPA 2001b; personal communication with CDC personnel, February 2003). This national guideline is published as general guidance for women living in areas where site-specific information is not readily available. ATSDR's conclusions for local consumption concerns are based on site-specific sampling data and; therefore, are preferable to this generic guidance.

    As a note, all of the fish and shellfish collected from Vieques contained levels of total mercury below FDA's guidance for methylmercury in fish, shellfish, and crustacea (1 ppm). The maximum concentration of total mercury in fish was a third of this guidance (maximum = 0.33 ppm, see Table 8), while the maximum concentration in shellfish was even lower (maximum = 0.049 ppm, see Table 9).

28. Comment: The data obtained in relation to the Fiddler Crabs are different from those obtained by Dr. Massol (Ecological Assessment of Heavy Metals in Vieques, Puerto Rico). The Fiddler Crab is within the feeding chain of the Blue Crab, which is consumed by humans (the Vieques population).

    Response: Dr. Massol and ATSDR analyzed fiddler crabs differently. Dr. Massol sampled and analyzed the extremities (levers and legs) of the fiddler crabs separately from the body (shell and internal contents). ATSDR analyzed the whole fiddler crab in one analysis. Therefore, one would expect differences in the results. Additionally, ATSDR acknowledged that rinsing fiddler crabs of sand and dirt was inadvertently omitted prior to placing them in sample containers. The washing methods were not stated in the report by Dr. Massol.

    ATSDR agrees that fiddler crabs are eaten by the blue land crab, and chemicals present in the fiddler crabs may then be transferred to the human food chain when people eat blue land crabs. To address this exposure pathway, ATSDR collected and analyzed blue land crabs (a species known to be consumed by the residents) from Vieques and evaluated whether eating them would result in harmful health effects (see Section V); rather than sample fiddler crabs (a species that is not eaten by the residents of Vieques). ATSDR determined that it is safe to eat blue land crabs from Vieques.

29. Comment: The disposition of contaminant products is not similar; it is necessary to enhance the research methodology from an epidemiological perspective.

    Response: ATSDR worked with the EPA/ERT to collect and analyze fish and shellfish from the coastal waters and near shore land on Vieques to determine whether fish and shellfish muscle tissues contain levels of heavy metals and explosives compounds that would adversely affect public health. During the July 2001 sampling, ATSDR and EPA/ERT collected 104 fish and 38 shellfish, representing 17 different species of fish and three different species of shellfish (see Table 7). The types of fish and shellfish collected were previously identified through several sources as species that are commonly caught and consumed by the residents of Vieques. This methodology is in accordance with EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories (EPA 2000).

30. Comment: Different environmental characteristics could affect the disposition and deposit of contaminants.

    Response: ATSDR agrees that, for a variety of reasons, different areas around Vieques may contain varying levels of chemicals. To evaluate whether eating fish and shellfish from different locations around Vieques would adversely affect public health, ATSDR and EPA/ERT caught fish and shellfish from five "preferred" fishing locations around Vieques--north of the LIA, south of the LIA, south of Esperanza, north of Isabel Segunda, and west of the former NASD--as well as from a local fish market in Isabel Segunda (see Figure 3). ATSDR determined that it is safe to eat fish and shellfish from all of the areas that ATSDR sampled. Even though there are differences in fish and shellfish body burdens between locations, these differences are too small to have an impact on public health.

31. Comment: Dr. Massol detected that the plant "Syringodium filiforme", which is consumed by the Manatee, was contaminated with lead. Therefore sea life contamination is present.

    Response: ATSDR's evaluation is a public health assessment, not an ecological assessment. Therefore, ATSDR's focus is on seafood that is eaten by the residents of Vieques. Fish and shellfish collected by ATSDR and EPA/ERT in July 2001 contained heavy metals, including lead, in their tissues (see Tables 8 and 9). This is not uncommon--heavy metals are commonly detected in fish and shellfish tissue because seafood tends to accumulate metals that are naturally present in the environment. However, all of the concentrations were detected at levels too low to present a public health concern to people who eat fish and shellfish from Vieques.

32. Comment: There has been no evaluation of specimens during exercises. In this case, the sea life could be greatly affected and enter to the human pathway.

    Response: Because ATSDR's involvement began in May 1999, after the Navy ceased military training with live munition, it was not possible for fish and shellfish to be collected during live bombing exercises. From July 16-20, 2001, ATSDR and EPA/ERT collected commonly consumed fish and shellfish from six locations on Vieques. The Navy conducted training exercises with inert ordnance from June 12-29, 2001, less than three weeks before ATSDR and EPA/ERT's sampling (personal communication with Navy personnel, February 2003). In addition, if fish contained higher levels of metals during a bombing exercise, the cumulative presence of the metals would have been detected during ATSDR and EPA/ERT's sampling.

33. Comment: The fish malformations are not evaluated.

    Response: Only two of 104 fish collected by ATSDR and EPA/ERT had any obvious deformities. One french grunt (Haemulon flavolineatum) had a deformed anal fin and a graysby (Epinephelus cruentatus) had a sunken belly. While the causes of these two malformations were not determined, it is not at all alarming or indicative of an unhealthy system to see so few fish with deformities.

34. Comment: I am not in accordance with the ATSDR conclusions, the methodology of the research must be changed in order to research for the proposed objectives.

    Response: ATSDR believes that the sampling and analyses conducted for this report completely address the important issue of whether contaminants are present in commonly consumed species of fish and shellfish at levels of health concern. During the July 2001 sampling, ATSDR and EPA/ERT collected 104 fish and 38 shellfish, representing 17 different species of fish and three different species of shellfish (see Table 7). The types of fish and shellfish collected were previously identified through several sources as species that are commonly caught and consumed by the residents of Vieques. This methodology is in accordance with EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories (EPA 2000). ATSDR concluded that it is safe to eat fish and shellfish from Vieques using standard public health evaluation procedures, which are described in greater detail in Appendix D.

35. Comment: Were samples collected at other times of the year to rule out if African dust storms had any influence on the presence of heavy metals detected?

    Response: It is important to look at the worst-case scenario in fish to be protective of public health. Despite being collected during a time when African dust storms could potentially cause an increase in the amount of chemicals around Vieques, the fish and shellfish are safe to eat (i.e., the concentrations detected in the fish and shellfish were not at a level of public health concern).

36. Comment: It should be clearly stated that the agricultural pollution resulting in water contamination is NOT related to Navy activities.

    Response: ATSDR acknowledges that the Navy does not engage in agricultural practices on Vieques.

37. Comment: Add to Table 2 a column titled "Frequency of Detection" to be consistent with Tables 3, 4, 5, 6, 7, and 8. It should be noted how many crabs out of approximately 140 crabs were detected to contain of the constituents noted.

    Response: ATSDR cannot add a frequency of detection column to Table 2, since the report produced by Casa Pueblo de Adjuntas and the University of Puerto Rico did not identify how many crabs contained each heavy metal.

38. Comment: What are "high concentrations" in reference to?

    Response: "High concentrations" was a term used by the authors of the report by Universidad Metropolitana (Caro et al. 2000) to describe the levels of arsenic, mercury, selenium, and zinc found in the fish they sampled. As explained in the text box, ATSDR determined that the levels of the metals detected in the fish are not high when compared to health-based values.

39. Comment: It can be concluded that Navy activities do not impact the mercury concentration levels in fish.

    Response: Based on ATSDR's evaluation, no activities on Vieques, including Navy activities, have resulted in levels of chemicals that would present a public health hazard to people eating the fish and shellfish.

40. Comment: Chain of custody forms and analytical results should be included as appendices.

    Response: ATSDR does not typically include chain of custody forms in appendices of PHAs. The analytical results, however, are summarized in the PHA. Chain of custody forms and analytical results are included in the Field Data Summary, Vieques Fish Assessment (EPA/ERT 2001), which establishes that the data used during ATSDR's public health assessment meet established EPA standards.

41. Comment: ATSDR should release all data in both tabular and electronic form identified by fish species, chemical species, date, sample location, detection limits and detection methods for both detects and non-detects. The Agency should also include its calculations for all exposure estimates that provide a basis for its claims of "no adverse health effects."

    Response: For clarity of presentation, within the PHA, ATSDR summarizes the analytical data generated during ATSDR and EPA/ERT's sampling and analysis activities. Table 8 summarizes the analytical results of the fish analyses and Table 9 summarizes the analytical results of the shellfish analyses. The fish species, chemical species, locations, detection limits, dates of collection, and methods are thoroughly discussed in this public health assessment.

    The calculations that ATSDR used to estimate exposures and to determine that it is safe to eat fish and shellfish from Vieques are included throughout Appendix D. The basic exposure dose equation and assumptions are provided on page D-1, as well as at the end of each of the tables within Appendix D.

42. Comment: I have reviewed the ATSDR fish and shellfish evaluation concerning Vieques Island. I remain concerned about the public health risks involving the residents of Vieques Island. While the ATSDR reaches the conclusion that there is no substantial risk from consuming fish and shellfish caught on and around Vieques Island, their sampling was not satisfactory. For example, the ATSDR purchased one cowfish from a fish market. This is problematic since, as the ATSDR stated in their report, "the cowfish contained the highest concentrations of arsenic (29.3 ppm) and selenium (2.5 ppm)." Cowfish is used as a food source as a filling in pastelillos.

    Response: The circumstances surrounding the collection of the one cowfish sample is explained in the Community Health Concerns section (Section VI). While conducting the sampling on Vieques, ATSDR met with the petitioner, who at that time, specifically requested that ATSDR collect and analyze boxfish (family Ostraciidae; e.g., cowfish and trunkfish). Realizing that the results and conclusions would be limited, ATSDR agreed to collect a representative sample from the fish market. Using the data from this limited sampling, ATSDR determined that consuming this cowfish and any others that contain similar levels of metals would not be expected to result in harmful health effects. Even though this one sample contained the highest concentrations of arsenic and selenium, it is not unusual for different samples of the same species to contain varying chemical concentrations--other cowfish samples may contain higher or lower levels of these chemicals. In addition, because of the specific uses of cowfish, this one fish would not be the only source of fish in a person's diet (i.e., other fish and non-seafood items would be eaten).

    During the July 2001 sampling, ATSDR and EPA/ERT collected 104 fish and 38 shellfish, representing 17 different species of fish and three different species of shellfish (see Table 7). The types of fish and shellfish collected were previously identified through several sources as species that are commonly caught and consumed by the residents of Vieques. This methodology is in accordance with EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories (EPA 2000).

43. Comment: The other area of concern is the presence of contaminates in fiddler crabs that the ATSDR would consider a risk if used for human consumption. However, the ATSDR does not consider fiddler crab contamination a serious public health concern since it is their understanding that the inhabitants of Vieques do not eat fiddler crabs. This is problematic for the following reasons.

    Response: As fiddler crabs are not a species that is consumed by the residents of Vieques, ATSDR did not evaluate them. Therefore, nowhere in the report does ATSDR identify that the presence of contaminants in fiddler crabs would be a risk if used for human consumption.

    a. Fiddler crabs may be a food source for other animals on Vieques Island that may form part of the food chain leading to ultimate human consumption. For example, birds may eat the fiddler crabs and humans may then eat the birds or their eggs. This type of analysis was not conducted in this report.

    Response: In discussions with the Fish and Wildlife Service (FWS), ATSDR was informed that herons and egrets eat fiddler crabs. However, many of these species are considered a Department of Interior trust resource and are protected under the Migratory Bird Treaty Act and Puerto Rican law and are; therefore, not likely to be consumed (personal communication with Fish and Wildlife Service personnel, February 2003).

    ATSDR acknowledges that fiddler crabs are eaten by the blue land crab, and chemicals present in the fiddler crabs may then be transferred to the human food chain when people eat blue land crabs. To address this exposure pathway, ATSDR collected and analyzed blue land crabs (a species known to be consumed by the residents) from Vieques and evaluated whether eating them would result in harmful health effects (see Section V); rather than sample fiddler crabs (a species that is not eaten by the residents of Vieques).

    b. The presence of contamination in the fiddler crabs makes it very likely that other food sources including fish and shellfish are contaminated. The ATSDR does not explain why fiddler crabs are the only animal life suffering from contamination on or around Vieques Island. If fiddler crabs are contaminated, it is difficult to see how the ATSDR can responsibly conclude that there is not risk to human health from daily consumption of fish or shellfish given the limited extent of testing conducted by the ATSDR.

    Response: Fiddler crabs are not the only species in which chemicals were detected. Other fish and shellfish species collected by ATSDR and EPA/ERT in July 2001 also contained heavy metals (see Table 8 and 9). This is not uncommon--heavy metals are commonly detected in fish and shellfish tissue because seafood tends to accumulate metals that are naturally present in the environment. However, all of the concentrations were detected at levels too low to present a public health concern to people who eat fish and shellfish from Vieques.

44. Comment: I do not feel comfort from the ATSDR report and would not personally consume indigenous food products from Vieques. I may have greater faith in the report if the personnel that drafted this report agreed to be stationed with their families, including minor children, on Vieques and ate an indigenous diet including daily consumption of local fish and shellfish.

    Response: ATSDR took great care to ensure that an objective evaluation was conducted to determine that the fish and shellfish are safe to eat. ATSDR's evaluation of whether fish and shellfish muscle tissues contain levels of heavy metals and explosives compounds that would adversely affect public health is conducted independent of the resident location of the researchers.

45. Comment: ATSDR compares the levels of methlymercury in fish from Vieques with the average level of methylmercury in canned tuna. ATSDR does not mention the concerns associated with consuming methylmercury in canned tuna, particularly for pregnant women, women of childbearing age, and children. The statement, "People who eat fish from Vieques would receive about as much mercury as people who eat canned tuna" implies that consumption of tuna and mercury contaminants is insignificant.

    Response: Information about the amount of mercury in canned tuna is supplied in the health assessment to provide perspective to the levels of mercury detected in Vieques fish and shellfish. According to FDA, consumption advice is unnecessary for canned tuna, since the methylmercury levels are less than 0.2 ppm (which is comparable to the average levels of total mercury in Vieques fish and shellfish) (FDA 2000). In addition, the average level of mercury in canned tuna (0.17 ppm) is a fraction of FDA's guidance for methylmercury in fish, shellfish, and crustacea (1 ppm).

46. Comment: The ATSDR does not address the potential for Vieques residents to exceed safe levels of exposure to contaminants such as methylmercury in fish caught nearby in addition to other sources such as canned tunafish. ATSDR should explain why it believes that pregnant women and children are safe from typically detected levels of methylmercury in tuna, in addition to mercury detected in Vieques fish. Cumulative exposure should be addressed for other contaminants released by the U.S. military on the island.

    Response: ATSDR's evaluation determines whether it is safe to eat fish and shellfish from Vieques. The information about canned tuna is supplied in the health assessment to provide perspective on the levels of mercury detected in Vieques.

    According to FDA, consumption advice is unnecessary for seafood species that have methylmercury levels less than 0.2 ppm (FDA 2000). The average mercury levels in fish and shellfish from Vieques and canned tuna are below this level (Vieques fish = 0.12 ppm, Vieques shellfish = 0.031 ppm, canned tuna = 0.17 ppm). In addition, the average levels of mercury in fish and shellfish from Vieques, as well as canned tuna, are a fraction of FDA's guidance for methylmercury in fish, shellfish, and crustacea (1 ppm).

    ATSDR's evaluation of eating fish and shellfish with the detected levels of mercury specifically takes into consideration pregnant women and women of child-bearing age. The health effects level used to determine that one could safely eat fish and shellfish from Vieques is based on the Seychelles Child Development Study, where no health effects were observed in children born to women who ate fish while pregnant and nursing (Davidson et al. 1998 as cited in ATSDR 1999a).

    Several studies, including those conducted by the National Toxicology Program in the United States and the TNO (Netherlands Organization for Applied Scientific Research) Nutrition and Food Research Institute in the Netherlands, among others, generally support the conclusion that if each individual chemical is at a concentration not likely to produce harmful health effects (as is the case on Vieques ), exposures to multiple chemicals are also not expected to be of health concern (for reviews, see Seed et al. 1995; Feron et al. 1993).

47. Comment: The Agency does not include a discussion of the sources and magnitude of uncertainty--including error--surrounding estimates of exposure. Nor does the Agency present quantitative estimates of uncertainty and statistical significance. The sources and magnitudes of uncertainty are so high in this case that the Agency should use Monte Carlo simulation techniques that employ full distributions of detected values and ranges of possible food intake to estimate distributions of exposure. This would allow the Agency to identify the proportion of the population likely to be exposed beyond the RfD.

    Response: The purpose of ATSDR's sampling and analysis activities was to determine whether the muscle tissues from commonly consumed fish and shellfish contain levels of heavy metals and explosives compounds that would adversely affect public health. The purpose was not to conduct a statistical evaluation. Therefore, ATSDR's analyses were sufficient for the public health evaluation that was conducted.

    To evaluate potential health hazards, ATSDR compared the oral health guideline values to exposure doses that were calculated using the maximum concentrations. A Monte Carlo simulation would not change this part of ATSDR's evaluation. If health guidelines were exceeded, ATSDR examined the health effects levels discussed in the scientific literature and more fully reviewed exposure potential, accounting for uncertainty in the assumptions used. During this part of the evaluation, ATSDR calculated exposure doses using average concentrations that were calculated using detected values only. To account for variability in the amount of fish people consumed, ATSDR calculated exposure doses for both adults and children who eat fish 7 times a week, 5 times a week, 4 times a week, 2 times a week, and once a week. Using this technique to account for uncertainty and variability is more transparent to the reader than a Monte Carlo simulation.

    Exposure to a dose above the MRL or RfD does not mean that adverse health effects will occur. MRLs and RfDs are intended only to serve as screening tools to help public health professionals decide whether to conduct a more comprehensive assessment of potential health effects. To maximize human health protection, MRLs and RfDs have built in uncertainty or safety factors (e.g., to account for human variability), making these values considerably lower than levels at which health effects have been observed. The result is that even if an exposure dose is higher than the MRL or RfD, it does not necessarily follow that harmful health effects will occur. This concept is further explained in the Using exposure doses to evaluate potential health hazards section in Appendix D.

    ATSDR further evaluated exposures to the chemicals detected above the MRL or RfD for both children and adults. All of the estimated exposure doses were below health effects levels reported in the scientific literature. ATSDR concluded that it is safe to eat fish and shellfish from Vieques using standard public health evaluation procedures, which are described in greater detail in Appendix D.

48. Comment: The Agency has not yet established a convincing "control" by taking fish samples from relatively pristine and undisturbed waters. This should be supplemented by a literature review of chemicals detections in highly consumed species from areas with no known anthropogenic source of the chemicals in question. The absence of systematic comparison species by species, and chemical by chemical, makes it more difficult to support the ATSDR "no adverse health effects" conclusion.

    Response: ATSDR's evaluation of Vieques fish and shellfish is a public health assessment, not an ecological assessment. The purpose was to determine whether the muscle tissues from commonly consumed fish and shellfish, collected from the coastal waters and near shore land on Vieques, contain levels of heavy metals and explosives compounds that would adversely affect public health. Therefore, the levels detected in the fish and shellfish were compared to health-based guidelines. Knowing whether the concentrations in Vieques fish and shellfish are higher or lower than other "pristine" areas would not change the conclusion that eating Vieques fish and shellfish would not result in adverse health effects.

49. Comment: ATSDR should clearly distinguish between the NOELs and NOAELs it relies upon to establish RfDs or MRLs, and all safety factors employed. In 1996, following recommendations of the National Academy of Sciences report, Pesticides in the Diets of Infants and Children, Congress enacted the Food Quality Protection Act to limit childrens' exposure to pesticides. This statute required the use of an additional 10 fold safety factor to account for the unusually high exposure of infants and children to pesticides, and their elevated susceptibility, related to their rapid rates of growth and development, and immature detoxification capacity. Given ATSDR's policy statement at the outset of their report, "As a policy, unless data are available to suggest otherwise, ATSDR considers children to be more sensitive and vulnerable to hazardous substances." (ATSDR 2003, Foreword), why would the Agency not adopt the more health protective safety factor suggested by NAS for pesticides? The level of food contamination and associated exposure levels are uncertain enough to justify a higher margin of safety.

    Response: Proposed MRLs undergo a rigorous review process: Health Effects/MRL Workgroup reviews within ATSDR's Division of Toxicology, expert panel peer reviews, and agency wide MRL Workgroup reviews, with participation from other federal agencies and comments from the public.

    MRLs are derived for hazardous substances using the NOAEL/uncertainty factor approach. They are below levels that might cause adverse health effects in the people most sensitive to such chemical-induced effects. MRLs are derived for acute (1-14 days), intermediate (15-364 days), and chronic (365 days and longer) durations and for the oral and inhalation routes of exposure. MRLs are generally based on the most sensitive chemical-induced end point considered to be of relevance to humans. Exposure to a level above the MRL does not mean that adverse health effects will occur.

    MRLs are intended only to serve as screening tools to help public health professionals decide whether to conduct a more comprehensive assessment of potential health effects. They may also be viewed as a mechanism to identify those hazardous waste sites that are not expected to cause adverse health effects. Most MRLs contain a degree of uncertainty because of the lack of precise toxicological information on the people who might be most sensitive (e.g., children, pregnant women) to the effects of hazardous substances. ATSDR uses a conservative (i.e., protective) approach to address this uncertainty consistent with the public health principle of prevention. Although human data are preferred, MRLs often must be based on animal studies because relevant human studies are lacking. In the absence of evidence to the contrary, ATSDR assumes that humans are more sensitive to the effects of a hazardous substance than animals and that certain persons may be particularly sensitive. Thus, the resulting MRL may be as much as a hundredfold below levels that have been shown to be nontoxic in laboratory animals.

    Specific details about the derivation of each chemical-specific MRL are provided in Appendix A of ATSDR's Toxicological Profiles. These Toxicological Profiles are available from the following URL: http://www.atsdr.cdc.gov/toxpro2.html.

50. Comment: Regarding the conch samples, it seems they were weighed with the shell on , did they weigh the animal once extracted from the shell? If not, this may skew your data since the shell is a considerable weight when compared to the animal tissue that is consumed by the public.

    Response: Yes, conch tissue samples were weighed without the shell for the chemical analyses. The laboratory analyzed one-gram tissue samples for the heavy metal analysis and two-gram tissue samples for the explosive compound analysis.

51. Comment: The Public Health Assessment Conclusions seem to be directed at a healthy population. It does not take into account persons consuming these food items that may already have debilitating diseases such as cancer or are undergoing medical treatment. A separate analysis should be done to see if this segment of the population could be at risk from regular consumption of fish and shellfish.

    Response: ATSDR does consider certain sensitive populations in the public health evaluation (e.g., children, pregnant women). However, this evaluation is intended for the general population of Vieques and does not address individual health concerns. If a person has specific concerns about their own personal health needs, the person should consult the local health department or a private physician.

52. Comment: Given the political ramifications of the desired departure of the Navy/Marine Corps from the island I think the basis for the petition is somewhat suspect; however, that is only supposition on my part.

    Response: Thank you for the comment.

53. Comment: Since the Navy is in fact stopping all activity on the Isla de Vieques in May, your study is coming out at a good time. Suggest it be in final form and promulgated as soon as possible.

    Response: Thank you for the comment.

54. Comment: The "Agency for Toxic Substances and Disease Registry" (ATSDR) presents the "Public Health Assessment" in a coordinated manner.

    Response: Thank you for the comment.

¹² EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories suggests collecting one species of fish and one species of shellfish when trying to assess whether a potential human health risk exists for estuarine/marine environments (EPA 2000).

¹³ EPA's Guidance for Assessing Chemical Contaminant Data for Use in Fish Advisories suggests collecting one species of fish and one species of shellfish when trying to assess whether a potential human health risk exists for estuarine/marine environments (EPA 2000).

¹⁴ Median total mercury concentration in 350 fish sampled from 25 species consumed by the Seychellois was <1 ppm (range, 0.004-0.75 ppm).