ENVIRONMENTAL JUSTICE:

A CASE STUDY OF COMMUNITY DIRECTED ASSESSMENT OF MERCURY IN FRESHWATER FISH IN WESTERN ALASKA

Lawrence K. Duffy, Tauni Rodgers and Senka Paul

Institute of Arctic Biology and Department of Chemistry and Biochemistry, University of Alaska Fairbanks, Fairbanks, AK 99775 (fychem@uaf.edu)

ABSTRACT

Increased awareness by subsistence users to the health risks associated with local mining activity and natural deposits of cinnabar prompted the Yukon-Kuskokwim Health Corporation (YKHC) to implement a mercury survey. People in the area served by YKHC are dependent on these fish for a large portion of their subsistence diet. There was concern and questions that mining activity and natural geological deposits in the area may lead to elevated levels of mercury (Hg) in these fish. Because of the actual and perceived health risk to humans, baseline data to assess such health risk was needed. Lush, Sheefish, Pike, Whitefish, Dolly Varden, Grayling, and Suckerfish were collected from various locations on the Yukon-Kuskokwim Delta. Frozen fish were transported to the laboratory, dissected, and tissues sent to an outside lab for mercury analysis using cold vapor atomic fluorescence. One hundred and eleven fish samples were collected from 18 locations. The Hg mean for all fish samples was 0.29 ppm. Ten fish were above the Food and Drug Administration criteria. The community was kept informed by radio programs and newspaper articles. However, a survey study could not address all of the questions raised in community forums. Our study supports the role of the community in initial research design and has provided valuable background data for comparisons with other areas in Alaska and Western Canada. This research raises questions about the role of glacial rivers and mercury transport in the Arctic.

INTRODUCTION

Metals appear in all arctic ecosystems and high concentrations of metals can have obvious impacts on the local environment. Plants and animals, including humans, usually have time to adapt over many generations. When natural erosion and transport of metals occurs as natural sources change with time, populations can continue to change because, on a relative time scale, these physical processes are slow enough to allow compensation by biochemical adaptation or behavioral modification. Anthropogenic pollutants associated with global climate change will have a major impact on local environments by disturbing the natural biogeochemical cycle. Moreover, impacts from human activities usually occur over a much shorter time scale, and biological adaptation may not be able to keep pace with these rapid environmental changes. Thus, a rapid increase in toxic metals can threaten the physical health of both plant and animal species, as well as human populations, that depend on these species for their subsistence.

Total mercury concentrations (Hg) in fish tissues are of special concern because of the potential of methylmercury to biomagnify through the food chain in aquatic ecosystems. Mercury, as methylmereury (MeHg) in fish, represents a potential risk to wildlife consumers such as piscivorous birds (e.g. guillemots and gulls) and mammals (e.g. mink and river otters) and possibly to the fish themselves (Braune et al., 1999). In Western Alaskan rivers, various species of fish, including northern pike (Esox lucius), burbot (Lota Lota), whitefish (Coregouus spp), grayling (Thymallus arcticus), and sheefish (Stenodus lencichthys), are important for subsistence among rural populations (Duffy et al., 1999). Mercury enters the Alaskan environment in two ways: (1) global distribution of industrial emissions through the atmosphere, and (2) from point sources, such as old mining areas and natural erosion of geological deposits. In the Yukon-Kuskokwim Delta, there was community concern that mining activity earlier in this century may have led to high mercury levels in local subsistence fisheries.

Several studies have demonstrated the presence of mercury in Alaskan subsistence users (Galster, 1976). Within the human population, cognitive defects in children with low level prenatal exposure to methylmercury has been reported (Grandjean et al., 1997). Because of the human health effects and ecological implications, a survey of mercury concentrations in fish muscle was begun in 1997 in Alaska. In this report, we compare our initial results with suggested critical values.

METHODS

Sampling: A combined total of 111 fish were sampled for mercury. The collection sites were distributed throughout the Yukon-Kuskokwim Delta region as chosen by subsistence users. Fish were collected by multiple collection methods. Efforts were made to have all fish samples donated by subsistence fishermen. Posters were faxed and mailed to village health aides requesting subsistence fish samples for the study. The posters gave a brief description of the study and why it was needed. Articles were also published in the Tundra Drums to inform the public and have fish samples donated. Posters were mailed to village schools and tribal councils to inform them of the proposed sampling and research project.

Fish samples were frozen when caught and remained frozen until they were dissected in the laboratory. Fish samples were thawed for two to three hours before dissection and cleaned of excess slime. Samples were weighed, measured, and placed on a clean dissection board. Muscle tissue was collected for total and methyl mercury testing by Frontier Geosciences Laboratory in Seattle, Washington. Fish samples were dissected using sterile surgical sheets, blades, and powder free latex gloves. Each fish sample was dissected with a new blade on a freshly cleaned dissection board.

Mercury Analysis: Hg was analyzed by cold vapor atomic fluorescence spectrophotometry (CVAF) after samples were digested with acid at Frontier Geosciences (Bloom and Fitzgerald, 1998).

Assessment: Regional assessment was conducted by comparing the means of individual species or river location sites with critical values. Critical values used in the analyses were those reported by Yeardley et al. (1998) as well as EPA, FDA and WHO standards cited in recent reviews (Braune et al., 1999; Van Oostdam. et al., 1999).

RESULTS

During 1997 and 1998. 111 fish samples were collected from 18 locations throughout the Yukon-Kuskokwim Delta region. CVAF spectrometry allowed Hg to be detected in all samples. The mean levels in the species collected showed that pike had the highest Hg mean of .533 ppm while dolly varden had the lowest mean. These differences reflect the trophic level at which these species are feeding. The mean level of Hg for all 111 fish was .287 ppm. There was no correlation between Hg levels and gender but mercury levels increased with the size of the fish. Methylmercury (MeHg) was the major form of mercury in muscle tissue. For a subset of 20 fish, the MeHg mean was .338 +/- .234 ppm while Hg mean was .334 +/- .487ppm.

Table I summarized the Hg concentrations of 18 locations in the Y-K Delta area. Andrefski, Bonsilla, Innoko and Piamute (near Holy Cross) showed the highest levels. The Andrefski River showed mean levels of 1.068 +/- 0.803 ppm, while Bonsilla was .565 +/- .234 ppm Hg. Lower levels occurred in fish from the Kuskokwim (near Bethel), the Johnson, and the Tuluksak. Holitna and Layman Lake were also higher than many other sites.

Since we know the Hg concentration in the Y-K delta fish, we can calculate a biomagnification ratio for fish consumers such as river otters or human subsistence users. We found the mean of 2 river otter hair samples from the Y-K Delta to be 2.7 ppm. Using our lowest level mean concentration in dolly varden and our highest mean level in pike, we calculated accumulation factors that ranged from 5 to 159 for river otters consuming either dolly varden or pike.

Similarly for human subsistence fishermen and women based on data for Hg in hair (Galster, 1976) the biomagnification factor using dolly varden is approximated 253 while for pike - human the biomagnification factor would be only 9. This is a 20 fold level of uncertainty that can only be overcome by accurate daily diet intake surveys which include identification of the type of fish consumed. Because mean hair levels of Hg from river otters are similar to the human mean in Alaska, river otters are a good sentinal species for a subsistence fishery.

Values of mercury in the Yukon River ranging form 6.5 to 28.7 ng/L depending on time of year and flow rate. These data would indicate a concentration factor for water to humans in the order of 10⁴.

DISCUSSION

Methylmereury (MeHg) is the species of mercury in the environment that has the greatest toxic effect in humans when there is chronic low level exposure. Microorganisms in the wetlands of the Y-K Delta are able to convert inorganic Hg to MeHg, so that it is transported up the food web (Branfireun et al., 1999). Also some bacteria in wetlands, such as those in the Y-K Delta may be incapable of utilizing C-1 compounds. MeHg is a C-1 compound and C-1 metabolic ratios are low in Alaskan wetlands so MeHg may be produced but not degraded (Hines and Morrison, 1992). High levels of MeHg can accumulate in fish muscle, such as pike, which in turn, becomes the primary environmental pathway of mercury to subsistance users in Alaska. MeHg in the human diet is almost completely absorbed into the blood stream and accumulates in muscle, kidney, liver and the central nervous system (CNS). At the exposure levels from fish reported in this study, the CNS would be the only target organ which may be adversely affected (Clarkson, 1992).

Further risk assessment which is based on scientific analysis for the Y-K Delta will require a evaluation of both the chronic MeHg exposure through fish consumption and its benefits. This risk weighing is difficult because of the lack of scientific information about health outcomes at

this approximate 3-10 ppm, in hair, exposure level (Van Oostdam et al., 1999). It is also hard to place a quantitative value on the nutritional and social benefits of subsistence style food consumption.

In risk characterization for the Y-K Delta, major uncertainties such as synergistic effects from light cycle, temperature. etc. may be involved, as well as possible covariables such as genetic/family history, education and other socio-economic factors. This difficulty in risk characterization is coupled with an insufficient number of options for subsistence users. Options such as regulatory measures or advisory measures are of little use since subsistence users do not have an economic alternative to high fish consumption. Also, it is difficult to conclude definitively whether or not a small intake exceedence, such as the Hg levels in the pike reported here, would constitute a real health risk. More research on mercury levels in fish, wildlife and humans as well as better data on CNS effects such as learning related to mercury burden will help resolve this risk management quandary.

One additional benefit of this study was the pragmatic integration of both scientific and traditional knowledge systems. Scientific risk assessment focuses on hypothesis testing by data collection and statistical analysis. Alaskan Native traditional knowledge is based on cumulative experience, close observation and oral knowledge communicated by elders and handed down over generations. Our study worked with the Y-K community leadership in initial research design, creating a collaborative environmental research project.

ACKNOWLEDGEMENTS

We thank Michelle Hecker for her assistance. This work was funded in part by the Cooperative

Institute for Arctic Research (CIFAR), a NOAA sponsored program, and a NIEHS pilot grant from the University of Washington’s Center for Ecogenetics.

REFERENCES

Bloom NS and Fitzgerald WE. (1998) Anal. Cheni. Acta, 208:151-159.

Branfireun BA, Roulet NT, et al. (1999) Global Biogeochem. Cycles. 13:743-750.

Braune B, Muir D. DeMarch B, et al. (1999) Sci. Total Environ, 230:145-208.

Clarkson TW. (1992). Environ. Health Perspect. I 00~3 1-38.

Duffy L.K., Scofield E, et al. (1999) Conip. Biochem. and Physiol., 24:184-186.

Galster WA. (1976). Environ. Health Perspect. 176: 15:135-140.

Grandjcan P. Weike P. White RF. et al.(1997) Neurotoxicol. Teratol. 19:417-428.

Hines ME, Morrison MC. (1992) J. Geophys. Res., 16:703-707.

Van Oostdam J, Gilman A, et al. (1999) a review. Sci. Total Environ., 230:1-82.

Yeardley RB, Lazorchak JM and Paulsen 8G. (1998) Environ. Toxicol. Chem, 17:1874-1884.

Table 1: Comparison of Mercury Concentrations at Different Locations in the Y-K Delta

Location	N	Mean (ppm)	Stand. Dev. (ppm)
Andrefski	3	1.068	0.802
Anvik	4	0.436	0.141
Bethel	21	0.228	0.432
Bonsilla	5	0.565	0.234
Emmonak	7	0.155	0.180
George	9	0.269	0.040
Goodnews	1	0.107	---
Gweek	1	0.178	---
Holitna	10	0.487	0.286
Innoko	2	0.521	0.165
Johnson	4	0.140	0.046
Kanektok	17	0.064	0.081
Kogrukluk	5	0.091	0.048
Kuskokuag	1	0.206	---
Kwethluk	5	0.158	0.089
Layman Lake	1	0.351	---
Paimute	11	0.522	0.404
Tuluksak	4	0.101	0.039