METHODS AND MATERIALS

Sample Sites

Sample sites represent areas that are downstream of former placer mining activities, and also non-mining areas for control and comparison purposes (Figure 1). A discrepancy between creek names was found on the USGS Nulato (B-4) Quadrangle map and a map used for a trapping pattern study by the Alaska Department of Fish and Game (ADF&G) Subsistence Division (Robert 1984). Bonanza Creek on the USGS Nulato (B-4) Quadrangle map originates in the Camp Creek drainage. It appears that Bonanza Creek is correctly labeled as Camp Creek on the ADF&G map. USFWS has referred to this sample site as Bonanza Creek, sample site 5 (Snyder-

Figure 1. Map of sample sites within the Kaiyuh Flats study area.

#3 Bishop Creek

#5 Camp Creek

#11 Bonanza Creek

#12 North Creek/American Creek

#13 Yukon Creek
Conn et al 1992b). Eddy Creek on the USGS Nulato (B-4) Quadrangle map is renamed as Bonanza Creek on the ADF&G map. North Creek is renamed American Creek on the ADF&G map at approximately Township 13S, Range 5E, and Section 4, SW 1/4. The ADF&G map, because it is consistent with the local knowledge and place names of the area, was used for this study. Description of sites in Figure 1 are as follows:

Site 3. Bishop Creek. USGS Nulato (C-3) Quadrangle. Township 9S, Range 8E, Section 35. Water samples collected. No fish captured in this drainage.

Site 5. Camp Creek. USGS Nulato (B-4) Quadrangle. Township 12S, Range 6E, Section 17, NE 1/4. Water and fish were collected immediately below the fork of Camp Creek in the NE 1/4 of Section 17.

Site 11. Bonanza Creek. USGS Nulato (B-4) Quadrangle. Township 12S, Range 5E, Section 35, NE 1/4. Water samples were collected in Section 35, NE 1/4. Fish were captured from Bonanza Creek within Sections 35 and 26.

Site 12. North Creek. USGS Nulato (B-4) Quadrangle. Township 13S, Range 5E, Section 9, NW 1/4. Water samples collected.

Site 12A. American Creek. USGS Nulato (B-4) Quadrangle. Township 12S, Range 5E, Section 30, SW 1/4. Fish captured.

Site 13. Yukon Creek. USGS Nulato (B-5) Quadrangle. Township 13S, Range 3E, Section 3, NW 1/4. Water samples and fish were collected within Yukon Creek immediately before the confluence with 22-Mile Slough (Khotol River).

All field work and sampling was conducted between 23-27 August, 1993. Of the five separate drainages sampled, Camp Creek has a history of gold placer mining, and Bishop Creek has a history of lead-silver production. North/American Creek, Bonanza Creek, and Yukon Creek have no documented mining history.

Permits applied for and received for this project included: 1) a conditional State of Alaska Department of Fish and Game Scientific Permit for collecting fish with a gill net, and 2) a conditional United States Department of the Interior Fish and Wildlife Service Refuge Special Use Permit.

A unique identification number was given to each fish captured and each water sample collected. The system used for this project is based on a system used by the USFWS within the Northern Unit of the Innoko National Wildlife Refuge. The identification number describes the area, sample site number, type of sample, sample number, and type of tissue. An example for a fish tissue sample would be as follows: NIN05PO01M, where NIN is the Northern Unit of the Innoko National Wildlife Refuge, 05 is the sample site, P is northern pike, 001 is sample number, and M is muscle tissue. An example for a surface water sample would be as follows: NIN 11 H2O 01, where NIN is the Northern Unit of the Innoko National Wildlife Refuge, 11 is the sample site, H20 is a surface water sample, and 01 is the sample number.

Surface Water

Surface water grab samples were collected at each site using 500 ml Nalgeneâ bottles just below the water surface and upstream of the collector. At the time surface water samples were collected, surface water temperature (° C) was recorded using a hand-held alcohol thermometer. pH was recorded using a Whatman model PHA300 portable pH meter. The meter was calibrated before each measurement with standard buffer solutions of pH 4.01 and pH 7.01. Water samples were chilled on blue ice and brought back from the field for analysis of turbidity, hardness, and conductivity. A duplicate surface water grab sample was collected at each site using a 500 ml plastic bottle provided by Columbia Analytical Services (CAS). The bottle contained nitric acid for the purpose of fixing the water sample to a pH of less than 2.0. The duplicate sample was shipped to CAS in for metals analyses.

Turbidity was measured in nephelometer turbidity units (NTU) using a Hach Portable Turbidimeter Model 2100P, calibrated with Hach Gelex secondary standards for 1, 10, and 100 NTU. Hardness was measured in parts per million (ppm) of calcium carbonate (CaCO3) using Hach hardness and digital titration methods and color endpoints. Conductivity was measured in umhos/cm using a Hach DREL/5â Conductivity meter with automatic temperature compensation. The USFWS Ecological Services, Fairbanks, provided the instruments for these measurements.

Fish

Fish were captured using conventional hook and line techniques. The target number of twelve (12) northern pike (Esox lucius) captured per sample site was achieved at four (4) of the five (5) sample sites for a total sample size of forty eight (48) specimens. Although effort was put forth, no fish were captured at Bishop Creek.

Total length (mm), fork length (mm), and weight (grams) were recorded for each fish prior to field dissection. Weight was measured using either a 2 kg, 5 kg, or 15 kg Chatillon hand-held scale. After fish metrics were recorded for each specimen, the same specimen was placed on a polyethylene dissection board that was pre rinsed with distilled water. Upon dissection, skin-off dorsal muscle tissue was removed from the right side of the specimen using a sterile stainless steel scalpel blade. The tissue sample was then placed into a new 125 ml plastic bottle and sealed. The corresponding identification number was recorded into a field notebook and written in permanent ink on the bottle. Each bottle was placed into a cooler with blue ice.

After each tissue sample was collected, the sex of each fish was recorded along with a cursory stomach content analysis. Scales and vertebrae were collected from each specimen for age analysis. Aging of vertebrae and scales was performed by the ADF&G Sportfish Division, Fairbanks, AK.

Prior to dissection of the next specimen, the dissection board was rinsed with distilled water and wiped dry with a disposable paper towel. A new sterile stainless steel scalpel blade was used for each tissue sample removed. A new pair of talc-free plastic surgical gloves was used when handling and dissecting each specimen.

Aware that livers, kidneys, intestines, and sex organs are routinely consumed by Kaiyuh Flats area residents, dorsal muscle was chosen for metals analysis due to the much greater amount of edible tissue provided per fish. The fact that individual body organs, such as the liver, may contain higher or lessor concentrations of contaminants is noted.

Upon return to Fairbanks, dry ice was placed in each cooler and the tissue and water samples were sent to Columbia Analytical Services (CAS) in Kelso, Washington for analysis of total mercury and total selenium.

Quality Assurance/ Quality Control (QA/QC)

Methods used by Columbia Analytical Services (CAS) for total mercury and selenium analyses were cold vapor atomic absorption, and graphite furnace atomic absorption, respectively. EPA methods used for analysis of total mercury and selenium in water samples were #7470 and #7740, respectively. EPA methods used for analysis of total mercury and selenium in fish tissues were #7471 and #7740, respectively. Total mercury analysis of surface water samples was completed within the required 28-day holding time (APHA et al. 1989).

Quality control screening used by CAS resulted in acceptable precision and accuracy for each metal analyzed in both fish tissue and water samples. The laboratory tested duplicate samples for relative percent difference (RPD), spiked duplicate samples for percent spike recovery, and analysis of standard reference materials (commercially provided tissue of known metal concentration). Therefore, the confidence level is high for the validity of the analysis results. The QA/QC laboratory report is presented in Appendix B.

RESULTS

Surface Water

Results of analysis for total metals concentration in surface water at each sample site were below the minimum limit of detection (LOD) of 0.0002 ppm for mercury, and 0.005 ppm for Selenium. Therefore, results of each sample analysis are reported as not detectable (ND) in Table 2.

For all sample sites, surface water temperatures ranged from 6.5° C to 11.0° C, (mean= 8.5° C). pH ranged from 6.85 to 8.08, (median= 7.68). Turbidity ranged from 0.31 NTU to 1.23 NTU (mean= 0.77 NTU). Conductivity ranged from 62 umhos/cm to 260 umhos/cm (mean= 141.4 umhos/cm). Hardness ranged from 27 ppm CaCO3 to 152 ppm CaCO3 (mean= 73.4 ppm). Water quality data is presented in Table 1.

Fish

Results of fish body size, age at capture, and metal concentrations of the muscle tissues collected are presented in Table 2. Fork length (FL) of fish captured ranged from 301 mm to 1010 mm (mean= 625). A length frequency of fish captured is presented in Table 3. All fish captured were sampled. Age of fish based on vertebrae aging ranged from 2 years to 19 years of age (mean= 6.8 years). Weight of fish ranged from 250 grams to 7000 grams (mean= 2130 grams). Twenty-five (25) males and twenty-three (23) females were captured resulting in a male to female ratio of 1.1:1.

All muscle tissue samples analyzed were above the LOD of 0.005 ppm for total mercury, but below the LOD of 0.400 ppm for total selenium. Total mercury in muscle tissue ranged from 0.091 ppm wet weight to 0.832 ppm wet weight (mean= 0.438 ppm). Columbia Analytical Services (CAS) analytical report is presented in Appendix A. CAS reported the metals concentrations in both wet weight (as received) and dry weight. The dry weight values can be converted to wet weight by using equation 1:

% Freeze Dried Solids

ppm Wet Weight 100 x ppm Dry Weight

Equation 1.

For all sample sites combined, concentrations of total mercury in muscle tissue show a strong positive correlation when plotted against the total length of the fish (Figure 2). The Pearson correlation coefficient is r=0.816.

Fish from the Camp Creek sample site, the creek where upstream gold placer mining had occurred, had a Pearson correlation coefficient of r=0.873 when mercury in fish tissue was plotted against the fork length of the fish. The remaining three drainages sampled, where no documented mining has taken place, resulted in the following Pearson correlation coefficients: North Creek/American Creek r=0.918, Bonanza Creek r=0.617, and Yukon Creek r=0.707.

For all sample sites combined, plotting the age of fish against the total mercury concentrations in fish tissues results in a strong positive correlation. The Pearson correlation coefficient is r=0.737. Total mercury plotted vs. age of fish is presented in Figure 3.

DISCUSSION

Water

Results from all water quality parameters sampled were within the Alaska Water Quality Standards 18 AAC 70. The low turbidity levels for Camp Creek and Bishop Creek suggest little or no physical water quality degradation, suspension, or resuspension of sediments, at least downstream to the sample sites was occurring. This may be directly related to the fact that no placer or lode mining activities were taking place during the sample period.

The absence of mercury and selenium being detected in surface waters is not necessarily surprising. Similar research has shown that while total mercury and selenium are below detection limits in surface waters, these same metals were above detection limits in fish muscle tissue (Speyer, 1980). The USFWS has data for two sites sampled within this study and also show mercury to be below detection limits in surface waters while fish tissues contained detectable amounts. The two sites sampled by USFWS in the study area were Bishop Creek and Camp Creek (Snyder-Conn et. al, 1992b).

Fish

All fish sampled, regardless of sample site, had mercury concentrations above method detection limits. This is consistent with preliminary results from a recent USFWS draft technical report for thirty-one (31) northern pike sampled at ten (10) sites within three national wildlife refuges in the area. For the Kaiyuh Flats total mercury wet weight muscle tissue concentrations reported for twelve (12) northern pike (assuming 25% freeze dried solids) ranged from 0.205 ppm to 0.563 ppm (mean= 0.403 ppm). Kidney and liver tissue also were analyzed for each specimen (Mueller et al. 1995).

Each individual sample site within this project has a strong positive correlation when Hg is plotted vs. fork length of fish and age of fish. These strong positive correlations are indicative of bioaccumulation (uptake via gills and skin) and biomagnification (food chain transfer) of mercury within the aquatic system. No specific sample site within the Kaiyuh Flats appeared to have fish with a significantly higher amount of mercury when correlated to fish size and fish age. Rather, mercury concentrations in fish muscle tissue appear to be ubiquitous.

The possibility of mercury accumulation as a result of off-refuge movements by northern pike was initially considered. During 1994 and 1995 the USFWS Fishery Resource Office, Fairbanks, AK and the ADF&G, Sportfish Division, Fairbanks, AK monitored northern pike movements in the Kaiyuh Flats using radio telemetry techniques. No off-refuge movements of northern pike were observed (Lubinski, USFWS, pers. comm.). It is unlikely that any off-refuge movements by northern pike are a contributing factor to on-refuge fish tissue mercury concentrations. Although, additional information on prey (whitefish) movements would assist in determining contributions to the mercury body burden of on-refuge northern pike.

Erosion of natural mercury deposits may be a source of mercury to the Kaiyuh Flats, but it appears that there are no documented cinnabar (HgS) deposits that could weather and drain into the Kaiyuh Flats catchment basin (Cruz and Cobb, 1984). Therefore, atmospheric deposition of mercury should be considered as a possible source of mercury to the Kaiyuh Flats.

Selenium concentrations in pike tissue were all below the LOD of 0.400 ppm wet weight. These results suggest that selenium may be present in insufficient quantities within the aquatic and biotic systems to potentially counteract the adverse effects of methylmercury. However, Mueller et al. (1995) reports selenium to be above detection limits of 0.500 ppm (reported as dry weight) in muscle, tissue, liver tissue, and whole body samples of all pike captured within the Kaiyuh Flats. After conversion to wet weight, the mean concentration for selenium in muscle tissue is 0.254 ppm wet weight, and the mean concentrations of mercury and selenium appear to be in a 1.6:1 ratio. The reason for this discrepancy is not known at this time, but could be a function of the relatively high method reporting limit of 0.400 ppm as reported by CAS. One researcher offered that the majority of selenium present in the specimens collected may have been concentrated in the liver and still provided an antagonistic effect on the mercury present in muscle tissue (P. Weber-Scannell, ADF&G, pers. comm. 1996).

Comparing the mercury concentrations in fish tissue from this project to guidelines used by the Minnesota Department of Health (Schubat 1990), the concentrations are considered to be low to moderate. However, all results of northern pike sampled for total mercury concentrations fell below the 1.0 ppm Food and Drug Administration (FDA) action level for consumption advisories.

Using the least squares method to calculate a straight line for mercury plotted against fork length results in the linear equation Y= -0.1662 + 0.00097 X. Mean length-at-age data from 925 northern pike captured from the Kaiyuh Flats during a USFWS/ADF&G 1994 radio telemetry project resulted in a fork length modal peak (n=161) at 642 mm and a maximum fork length of 1160 mm (Taube 1995). Substituting 642 mm and 1160 mm for X in the linear equation and solving for Y, results in calculated mercury concentrations of 0.457 ppm and 0.959 ppm, respectively. Therefore, based on the linear equation derived from this project, the largest specimen from a sample size of 1004 northern pike [48 fish from this project, 31 from Mueller, et al. (1995), and 925 from the USFWS/ADF&G 1994 project] does not exceed the 1.0 ppm FDA action level.

Fish Consumption and Risk Assessment

Several states have issued fish consumption advisories concerning specific species with elevated concentrations of mercury. At this time the State of Alaska has no advisories for fish consumption in marine or fresh waters. FDA action levels are essential on a national and inter-state basis, but may not be appropriate for local and site-specific situations where more definitive information is available.

An example of using results from this project in a fish consumption advisory model would be as follows: If the average size of fish that are harvested is 625 mm (È 25 inches), the calculated concentration of mercury in muscle tissue (based on the linear equation of Y= -.1662 + 0.00097 X) would be 0.440 ppm. Using this 0.440 ppm value in a model (Equation 2) proposed by Dourson and Clark (1990), the recommended amount of fish muscle tissue to be consumed over a year-long period without any adverse health effects would be approximately 17.4 kilograms (38.3 lbs). This advised amount of consumption is calculated for an adult male whose body weight is 70 kg (I 54 lbs.).

RfD (mg of chemical/kg bw/day) X 70 (kg bw)

Fish intake (kg of fish/day)= fish concentration (mg of chemical/kg of fish)

Equation 2. Where reference dose (RfD) for mercury is 3 E-4 and (bw) is body weight.

Schubat (1991) recommends that sensitive individuals (pregnant women, nursing mothers, and children under six (6) years of age) consume only 25% of this amount, therefore reducing the recommended amount of consumption to approximately 4.4 kg. (9.6 lbs.) per year. Table 3 presents various fish sizes, calculated mercury concentrations, and recommended consumption amounts based on Shubat (1991) and Dourson and Clark (1990).

It should be noted that the Minnesota Department of Public Health has classified fish consumers as either long term (chronic), 3-12 months/year, and short term (acute), 3 months/year or less. Long term more consistent exposure poses more of a concern because as the period of exposure increases the adverse effect concentration (ppm) decreases. Nearly all of the RfDs are based on long-term daily exposure, therefore, incorporating a more conservative approach into fish advisory models. Based on models proposed by Shubat (1991), it would be safe to chronically consume 2.8 meals/month of fish containing 1.0 ppm. In this case 1 meal would be considered 0.5 lbs.

The reader should be aware of the considerable debate and dissatisfaction among scientists, health care professionals, regulators, and ordinary citizens regarding human health risk assessment associated with mercury in the environment (Graham 1995). The fish consumption models described above use reference doses (RfD) within their respective equations. A RfD is an estimate of daily exposure of the human population to mercury that would not result in an appreciable risk of deleterious effects during a lifetime. However, new approaches are being explored using a "benchmark" approach to more accurately estimate RfDs. The "benchmark" approach utilizes all data points in a set rather than relying on individual data points for RfD estimation.

Depending on an individual's fish consumption habits, level of consumption, and body weight, the recommended consumption levels will vary. Due to the lack of published data on actual consumption rates for the residents of the villages of Kaltag, Nulato, and Koyukuk, estimates of consumption rates are variable and are based on personal communication.

Area residents continue to use the Kaiyuh Flats area for numerous subsistence activities. Northern pike are generally harvested using hook and line techniques. Pike are eaten on site and/or brought back to the villages for human consumption as well as dog food. Northern pike are prepared for consumption in various ways, including: head off and boiled, fried guts (cleaned intestines), livers, kidneys, and gonads, and boiled and made into "ice cream." "Ice cream" is a combination of boiled pike muscle tissue, moose fat, and berries. The mixture is frozen and usually eaten within two weeks. Although some area residents regularly harvest and consume northern pike, there are still concerns of elevated mercury concentrations in the fish. As a result, there is a general feeling that the larger fish should not be eaten. Additionally, it is difficult to quantify the amount of fish harvested and consumed on a yearly basis (G. Yaska and C. Peter, pers. comm. 1995).

Subsistence harvests of northern pike by Galena area residents during 1985 and 1986 have been reported by the ADF&G as 5.2 pounds of edible fish per capita (Marcotte 1990). This report does not describe the area of harvest, therefore, an unquantified amount of the northern pike may have been harvested from the Yukon River mainstem.

Possible comparisons may be made between the villages of Kaltag and Nulato to other villages in interior Alaska that are located in similar basin settings. The villages of Huslia and Minto are each situated in lake/river regions and residents fish for pike in open water and through the ice. Values for edible pounds of pike harvested per capita for Huslia was 28.8 pounds in 1983, and 76.7 pounds for Minto in 1984 as reported by the ADF&G (Walker et al. 1988). These reported values represent the edible pounds of pike harvested. The need for further research on actual pounds of fish consumed on a per capita basis must be stressed.

RECOMMENDATIONS

1) It is the recommendation of TCC's Wildlife & Parks Program that results from this project be used by the State of Alaska Department of Health and Social Services, Division of Public Health, to determine whether guidelines are needed concerning consumption rates of northern pike harvested within the study area. Concern remains regarding recommended consumption of fish by sensitive individuals (pregnant women, nursing mothers, and children under six (6) years of age).

2) Additional information should be gathered regarding actual amounts of fish consumed per capita by Kaiyuh Flats area residents.

3) Due to potential hard rock mining activities proposed for the Kaiyuh Mountains, periodic heavy metals information should be gathered from sediment, water, and fish tissue within the Kaiyuh Flats area.
Go to References