100
R.R. Turner and G.R. Southworth
true for the Wabigoon River alone. Mercury concentrations in biota actually
increased with distance downstream. This pattern was ascribed to increasing
bioavailability of mercury with distance. Thus, although water and sediment
mercury concentrations decreased in the downstream direction, the fraction that
was methylated and thus most bioavailable increased in downstream direction.
Data from this site support the hypothesis that general water-quality conditions
are far more important in regulating mercury in fish than total mercury in water
and sediment.
2.8
Saltville, Virginia
Chemical plants have operated along the banks of the North Fork of the Holston
River (NFHR) in Virginia since the late 1800s. Construction and operation of a
chloralkali plant from 1952 to 1972 resulted in Hg contamination of the river
downstream of the plant and of a large reservoir 100 km downstream. An
estimated 100 Mg of mercury was released into the river (NUS 1983). Disposal of
mercury waste (approx. 45 Mg; NUS 1983) into large settling ponds containing
solid Solvay wastes (from soda ash production) has resulted in extensive seepage
of mercury into the river long after the facility was closed and the chloralkali
plant building razed. Mercury contamination of fish from the North Fork
Holston River in Virginia was first discovered in the late 1960s (EPA 1995;
Derryberry 1972; Bailey 1974; Toole and Ruane 1976; Hildebrand et al. 1980a),
soon after mercury bioaccumulation became recognized as a concern associated
with chloralkali plant discharges. In 1970, in-plant controls at the facility reduced
mercury losses to approximately 100 g day -I. All discharges from ongoing
operations ceased with closure of the plant in 1972. Four years after the plant was
closed, mercury releases to the NFHR, mainly from pond seepage and overflow,
were estimated (Lindberg and Turner 1977; Turner and Lindberg 1978) to be
50 g day-I, with comparable air emissions from the surface of the waste ponds.
In 1982-1983 river sediments along approximately 600 m of stream were
dredged and the rock bottom grouted to seal in mercury which could not be
removed (TV A 1983) The mercury cell building was razed and the site capped
with clean fill, including placement of rip rap along the river bank adjacent to the
former plant site. Other actions were taken to reduce percolation of runoff and
direct rainfall through the large waste ponds which had been used to dispose of
aqueous wastes from the mercury cell operation. However, considerable (800
LPM) seepage water from this pond, with mercury concentrations around
50 ~lg 1-1 continued to be discharged into the adjacent river. This flow was finally
directed through an activated carbon-based treatment system in 1995.
Summertime baseflow measurements of total inorganic and methylmercury
(unfiltered samples) made by us on samples from NFHR near Saltville in 1996
and 1997 found water-borne inorganic mercury to be relatively low for a
contaminated site. Aqueous total mercury averaged 17 ± 6.5 ng 1-1 'on three
sampling dates. Methylmercury was 0.19 and 0.49 ng 1-1 when sampled in 1996
and 1997. Typical values for uncontaminated streams in geologically similar
R.R. Turner and G.R. Southworth
true for the Wabigoon River alone. Mercury concentrations in biota actually
increased with distance downstream. This pattern was ascribed to increasing
bioavailability of mercury with distance. Thus, although water and sediment
mercury concentrations decreased in the downstream direction, the fraction that
was methylated and thus most bioavailable increased in downstream direction.
Data from this site support the hypothesis that general water-quality conditions
are far more important in regulating mercury in fish than total mercury in water
and sediment.
2.8
Saltville, Virginia
Chemical plants have operated along the banks of the North Fork of the Holston
River (NFHR) in Virginia since the late 1800s. Construction and operation of a
chloralkali plant from 1952 to 1972 resulted in Hg contamination of the river
downstream of the plant and of a large reservoir 100 km downstream. An
estimated 100 Mg of mercury was released into the river (NUS 1983). Disposal of
mercury waste (approx. 45 Mg; NUS 1983) into large settling ponds containing
solid Solvay wastes (from soda ash production) has resulted in extensive seepage
of mercury into the river long after the facility was closed and the chloralkali
plant building razed. Mercury contamination of fish from the North Fork
Holston River in Virginia was first discovered in the late 1960s (EPA 1995;
Derryberry 1972; Bailey 1974; Toole and Ruane 1976; Hildebrand et al. 1980a),
soon after mercury bioaccumulation became recognized as a concern associated
with chloralkali plant discharges. In 1970, in-plant controls at the facility reduced
mercury losses to approximately 100 g day -I. All discharges from ongoing
operations ceased with closure of the plant in 1972. Four years after the plant was
closed, mercury releases to the NFHR, mainly from pond seepage and overflow,
were estimated (Lindberg and Turner 1977; Turner and Lindberg 1978) to be
50 g day-I, with comparable air emissions from the surface of the waste ponds.
In 1982-1983 river sediments along approximately 600 m of stream were
dredged and the rock bottom grouted to seal in mercury which could not be
removed (TV A 1983) The mercury cell building was razed and the site capped
with clean fill, including placement of rip rap along the river bank adjacent to the
former plant site. Other actions were taken to reduce percolation of runoff and
direct rainfall through the large waste ponds which had been used to dispose of
aqueous wastes from the mercury cell operation. However, considerable (800
LPM) seepage water from this pond, with mercury concentrations around
50 ~lg 1-1 continued to be discharged into the adjacent river. This flow was finally
directed through an activated carbon-based treatment system in 1995.
Summertime baseflow measurements of total inorganic and methylmercury
(unfiltered samples) made by us on samples from NFHR near Saltville in 1996
and 1997 found water-borne inorganic mercury to be relatively low for a
contaminated site. Aqueous total mercury averaged 17 ± 6.5 ng 1-1 'on three
sampling dates. Methylmercury was 0.19 and 0.49 ng 1-1 when sampled in 1996
and 1997. Typical values for uncontaminated streams in geologically similar