Polybrominated biphenyl (PBB) contamination of the Pine River, Gratiot, and Midland Counties, Michigan.

Michigan Chemical Corporation, St. Louis, Michigan manufactured PBB from 1970 until November 20, 1974. Studies in 1974 showed significant quantities of PBB in effluent discharged from the facility and in water, fish, ducks, and sediments from the Pine River. Fish uptake rates and bioconcentration factors were estimated. Followup surveys over the three year period since the termination of PBB production indicate a decline in PBB loadings to the river but no significant corresponding decline of PBB levels in sediments, fish and duck tissue. A Michigan Department of Public Health warning against consumption of Pine River fish from St. Louis downstream to its confluence with Chippewa River remains in effect.


Introduction
Michigan Chemical Corporation, St. Louis, Michigan was the major manufacturer of polybrominated biphenyls (PBB) in the United States. Production of their PBB flame retardants, FireMaster BP-6 and FireMaster FF-1, totalled approximately 11,323,000 lb between 1971 and 1974 ( Table 1). The company produced, bagged, and marketed the raw products to be used as a component of thermoplastics in a variety of applications. Production of PBB by Michigan Chemical Corporation was voluntarily stopped on November 20, 1977 upon request by the Michigan Department of Natural Resources (DNR).
Chemically, PBB is similar to the more commonly known contaminant polychlorinated biphenyl (PCB). The main structural difference is the attachment of bromine rather than chlorine on the biphenyl molecule. FireMaster BP-6 and FF-l are mixtures of brominated biphenyls with an average bromine content equivalent to six bromine atoms per molecule. They are somewhat less soluble in water (11 ppb at 25°C) than PCB. Unlike PCB, FireMaster BP-6 and FF-I are solids at room temperature. Since PBB is not used as a liquid, it is much less likely to contaminate the environment than is PCB. PBB is also reportedly not easily leached from the thermoplastics in which it has been incorporated. Therefore the potential for contamination of the aquatic environment by PBB appears to be primarily related to production and formulation processes rather than to usage. This report evaluates the contamination of the Pine River in Gratiot and Midland counties downstream from the Michigan Chemical Corporation plant. The Pine River is part of the Saginaw River system which flows into Saginaw Bay of Lake Huron. The Pine River flows through portions of Isabella, Gratiot, and Midland Counties and joins the Chippewa River 2 miles upstream from its confluence with the Tittabawassee River at Midland. Michigan Chemical Corporation, St. Louis, is located on the Pine River approximately 29 river miles upstream from the Pine's confluence with the Chippewa. The Pine River at St. Louis has a oncein-ten years, seven-day drought flow of about 27 ft 3/sec. Surveys were conducted from 1974 through 1977 to determine PBB concentrations in Michigan Chemical Corporation effluent discharged to the Pine River, river water and sediments, and resident fish and duck populations. PBB uptake by introduced fish held captive in the Pine River at various locations during active PBB production was also measured.

Methods
Samples of company effluent were obtained by either grab sampling or collection of a composite sample over a 24-hr period into 1-gal glass containers equipped with Teflon-lined caps and containing 100 ml of redistilled hexane as a solvent. River water samples were collected August 19-20, 1974 by a 30 hr glass composite sampler described by Zillich et al. (1) and on other occasions by grab sampling into 1-liter glass containers.
Sediments collected on July 22, 1974, were collected with an Ekman dredge. The surface layer was removed and transported to the laboratory in Whirlpak bags. Grab samples of the surface layer of sediment were hand collected into wide-mouth 250 ml glass bottles on all other sampling dates.
Resident fish populations were collected by angling, gillnetting, or chemical treatment of the river with rotenone. All fish samples were skinless fillets. Fathead minnows (Pimephales promelas) with an approximate average weight of 2 g each were used in the caged fish uptake study conducted October [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18]1974. The minnows were obtained from the Department of Natural Resources Wolf Lake Fish Hatchery. The fish cages were constructed of 3/4-in. plywood, 1 ft square and 1 ft deep. Screened openings on two sides allowed water circulation. Fifteen to eighteen minnows were placed in each cage. Two cages, anchored offshore to a steel fence post, were used at each of four stations. The fish were conditioned to the ambient river water temperature for 15 to 20 min prior to being placed in the cages. Samples of minnows (composites of 10-15 whole minnows) were collected at the time of introduction and after one-week and twoweek periods of exposure. The minnows were individually wiped with dry paper towels to remove the slime layer. The fish were then wrapped in aluminum foil and frozen, until analysis. One-gallon grab water samples were also collected from each station at the time of introduction and at the one-and two-week sampling times.
Ducks killed on the Pine River below St. Louis during October 9-16 were obtained from local hunters on January 8, 1974. The hunters had skinned, eviscerated, and frozen the ducks soon after shooting them. The ducks were returned to Lansing and thawed. Breast muscle was removed, packaged in aluminum foil, and refrozen. On September 28, 1976 ducks were collected by Department of Natural Resources staff from Section 18, T12N, R2W downstream from St. Louis. One side of the breast (muscle and skin only) of each duck was removed after all feathers had been plucked. The other side of the breast was removed after it had been skinned. The breast portions were then grouped into composite samples according to species and whether they were plucked or skinned. In 1977, ducks were again collected by Department of Natural Resources staff from Sections 17, 18, and 19, T12N, R2W. Paired samples of skinned and plucked breast tissue were analyzed.
Company effluent and river water and sediment samples were analyzed at the Michigan Bureau of Environmental Protection Laboratory in Lansing. Effluent and river water samples were extracted with redistilled hexane:ethyl ether (85:15) and analyzed for polybrominated biphenyl residues via electron-capture gas chromotography. The instrument used was a Varian Aerograph equipped with a Nickel-63 detector. A glass column, V8 in. x 6 ft packed with 5% OV-210 liquid phase on QF-1 (100-120 mesh) solid support was used for all analyses. Carrier gas was purified nitrogen at a flow rate of 28.0 ml/min. Column temperature was kept isothermal at 2250C and the detector was maintained at 315°C. Peak response was measured by a Varian millivolt recorder with a 0.1 mV signal producing full scale detection. The minimal detectable level using these conditions was 0.1 ,g/l. for a 1.0 liter water sample. Spiked samples gave an average recovery of 90.0%. Sediments were analyzed, as described above for water. Sediment extractions, however, were made using a hexane:acetone mixture at 2: 1 ratio.
Resident fish samples obtained on June 19, 1974 were analyzed by the Food and Drug Administration, Division of Chemical Technology Laboratory in Wa:hington, D. C. Resident fish sampled on all subsequent dates and all duck samples were analyzed by the Michigan Department of Agriculture Laboratory, East Lansing, Michigan. Analysis of both water and fish samples in the caged fish uptake study were performed in the Michigan DNR Bureau of Environmental Protection Laboratory in Lansing using methods comparable to those used for sediment analyses.

Michigan Chemical Corporation Discharges
Eight Michigan Chemical Corporation outfalls to the impoundment on the Pine River at St. Louis have been sampled for polybrominated biphenyl (as BP-6). The highest concentrations found in any outfall were detected by company sampling of outfall 005 in 1972. Values ranged from 98 to 503 ,g/l. The source of contamination was within the BP-6 production area where particles of PBB were lost to the cooling water during the flaking operation. To alleviate this problem, additional shielding was added to the flaker and a brine chiller was installed to recirculate the flaker cooling water. Condenser cooling water was repiped to outfall 006. Outfall 005 was then abandoned. Company officials believed all PBB losses to the river had been eliminated by these actions.
The Michigan Bureau of Environmental Protection began an investigation in June, 1974 following the disclosure that farm animals had been accidentally contaminated and poisoned by PBB (2). Outfall 006 was sampled first, since it was the only outfall from the BP-6 production area. The highest PBB concentration detected in the effluent from this outfall was 10.6 ,g/l.
It was not until September, 1974 that additional outfalls not directly associated with BP-6 production were sampled. Six additional outfalls (002, 003, 004, 008, 013, and 014) were found to contain measurable PBB concentrations with some far exceeding concentrations detected in outfall 006. Hydrobromic acid, produced as a by-product in the BP-6 operation and utilized in several other production areas, was thought to be the source of contamination in five of the six other contaminated outfalls. The low level PBB contamination in outfall 008 was thought to originate in intake water drawn from the impoundment a short distance from contaminated outfalls 013 and 014. Based upon flow rates and concentrations found for the seven contaminated outfalls, the quantity of PBB being discharged to the river during the September 4-5, 1974 survey was calculated to be 167 g/day ( Table 2).
Even with all PBB production and utilization of by-products halted, concentrations of PBB in the effluent discharged to the Pine River continued to be high on occasions. Although erratic, concentrations as high as 150 ,g/l. were detected following shutdown. Explanations for continued losses include release of PBB accumulated in the discharge pipes and inflltration of surface water which might accumulate PBB from contact with contaminated soils. Soil samples collected from the bagging and loading areas were found to contain 3500 and 2500 mg/kg PBB, respectively. These highly contaminated soils were removed upon order of the DNR and lesser contaminated areas were sealed with asphalt to protect against further infiltration of precipitation. Most recent effluent samples show negligible PBB levels. Calculated daily loadings decreased to less than 1 g/day by August 1977 (Table 2) Table 3.
Concentrations upstream from the Michigan Chemical Corporation plant were all less than the analytical sensitivity limit of 100 ,ug/kg with one ex-

Resident Fish Tissue Analyses
Elevated PBB levels were found in several Pine River fish species ranging up to a maximum of 1.33 mg/kg in carp (skinless fillets; wet weight) captured in the reservoir in the vicinity of Michigan Chemical Corporation (Table 4). Significant levels of PBB were present in fish from each station sampled during 1974 and 1976. No PBB was detected in fish from the Alma Reservoir which is upstream from Michigan Chemical Corporation and above a dam which prevents upstream fish movement. Samples of carp (Cyprimus carpio) and freshwater drum (Aplodinotus grunniens) collected and analyzed from the Tittabawassee River in 1974 did not have detectable concentrations of PBB. The extent of fish contamination appears to be limited to the Pine River downstream from St. Louis. There was no apparent change in PBB concentrations in fish between 1974 and 1976.
The U. S. Food and Drug Administration (FDA) has not established a tolerance limit for human consumption for PBB in fish. The tolerance limit for beef was originally established at 1.0 mg/kg (fat weight basis) and later lowered to 0.3 mg/kg. Fish with 1.33 mg/kg (wet weight basis) would contain approximately 30 mg/kg PBB on a fat weight basis. This is obviously greatly in excess of the FDA tolerance limit for beef. On November 21, 1974 the Michigan Department of Public Health issued a warning against consumption of fish from the contaminated zone of the Pine River. This warning was renewed following a review of the 1976 fish data.

Caged Fish Uptake Study
Uptake of PBB's by fathead minnows (Pimephales promelas) occurred at two of the test stations (Table 5). No PBB's were detected in either the fish or the water at Station 1. This station was about three miles upstream of the Michigan Chemical plant. At Station 3, all the fish died from undetermined causes before one week had passed. PBB residues in the water at this site ranged from 0.4 to 1.3 ,ug/l. At Station 2, a mean of 3.9 ,g/l. PBB was measured in the water. The range of three measurements was from 0.9 to 9.8 ,tg/l. Seven days after introduction, the fish accumulated 800 ug/kg measured on a wet weight basis. After two weeks exposure the fish had accumulated 1000 ,ug/kg PBB.
At Station 4, below the dam in St. Louis, the  levels of PBB in the water were below the level of analytical sensitivity (0.1 ug/l.). After 1 week the fish accumulated 200 ,ug/kg, and after 2 weeks had accumulated 1000 ,4g/kg.
The PBB uptake by fish held in cages indicates that extremely low levels of PBB in the water are sufficient to cause unacceptable concentrations in fish flesh in a very short period of time. At Station 4, where water concentrations remained consistently less than 0.1 ,ug/l., the caged fish accumulated 1000 ,ug/kg in their tissue. This represents a concentration factor of greater than 10,000-fold in just two weeks of exposure. Bioaccumulation of PBB's, therefore, appears comparable to polychlorinated biphenyls (PCB's). Hansen et al. (3) reported PCB concentration factors in two fish species to be 10,000 to 50,000 times the exposure levels in water. Other estimates of PCB accumulation ratios by fish vary from 26,000 to 200,000 times the exposure levels in water (4).

Ducks
Concentrations of PBB in the flesh of three wild duck species (mallards, teal, and wood duck) col-  Table 6. PBB in skinless breast tissue from composite samples of mallards, wood ducks, and teal in 1974 was 0.25, 0.29, and 1.8 mg/kg, respectively, on a whole weight basis. Samples were either not available for all species from year to year or not equal in number. Concentrations of PBB in samples with skin left on are considerably higher than in skinless samples indicating much of the PBB is associated with the skin or fat layer between the skin and muscle. Interpretation of residue data for wild ducks is complicated because of their migratory habits. Nevertheless, the majority of ducks analyzed from the Pine River contained measurable concentrations of PBB.
This study has been possible through the cooperative efforts of many individuals and agencies. Employees of the Environmental Service, Fisheries, Law, Water Quality, and Wildlife Divisions of the Michigan DNR were responsible for planning, collection of samples and analysis of data.