Page 817 - The Toxicology of Fishes

Mining Impacts on Fish in the Clark Fork River, Montana: A Field Ecotoxicology Case Study 797

Underwater observations (via snorkeling) and electrofishing were used to estimate trout numbers and
biomass in 100-m comparable segments. Removal–depletion electrofishing was used where poor water
clarity precluded the use of snorkeling, as well as for validating snorkel estimates. In 1991, reference
segments contained significantly more juvenile and adult trout (brown and rainbow trout) and more
biomass than Silverbow Creek and the upper Clark Fork (Figure 19.11). Reference trout densities
averaged 5.3 times greater than densities in contaminated segments. The Physical Habitat Simulation
Model developed by the U.S. Fish and Wildlife Service was used to address differences in habitat
(Hillman et al., 1995). This model calculated the weighted useable area (WUA) of trout habitat for a
given segment. Numbers and biomass were divided by the WUA to account for differences in habitat.
A comparison of adjusted trout populations in both Silver Bow and Clark Fork and reference segments
showed the same pattern: Adjusted adult and juvenile trout densities in reference segments were 5.8
times greater than densities in the Silver Bow and Clark Fork segments. Hillman and Chapman (1995)
repeated their studies in 1994 with similar results. Total trout densities in the Clark Fork segments ranged
from 7 to 188 trout per hectare. Trout populations in reference segments ranged from 39 to 528 trout
per hectare. In addition, although trout densities fluctuated over the course of the summer in each of the
four Clark Fork segments, similar patterns of change were observed in the reference segments, and the
difference between Clark Fork and reference segments remained significant at all times.
These careful comparisons come as close as possible to documenting that population differences are
not accounted for by differences in trout habitat. Overall, the population studies substantiate the uniquely
low abundance and diversity of trout in the Clark Fork River compared to streams affected by stressors
typical of Montana streams but unaffected by mine wastes.

Metal Contamination Is Bioavailable to Trout
Corresponding to the elevated concentrations of metals in water, sediment, and benthic macroinverte-
brates, the resident fish in the Clark Fork River also contain elevated tissue metals. Farag et al. (1995)
documented elevated arsenic, cadmium, copper, and lead in the gill, liver, kidney, pyloric ceca, stomach,
large intestine, stomach contents, and whole fish of brown trout from two sites in the Clark Fork,
compared to brown trout collected from two reference sites. Tissue concentrations of copper were greater
than 2300 µg/g (dry wt) in livers and greater than 1250 µg/g in the gill, pyloric ceca, and stomach tissues
of brown trout sampled near Warm Springs Ponds (Figure 19.12). Similar concentrations were measured
in April of 1989 (Phillips and Spoon, 1990) and in August and November of 1991. In earlier years (Dent,
1974), the concentrations were greater than the concentrations reported above. The elevated metal
concentrations in specific tissues of fish from the Clark Fork River indicated that resident fish in this
river system were exposed to bioavailable copper and acquired a tissue dose of metals in specific organs.
Farag et al. (1995, 1999) suggested that a liver concentration between 238 and 480 µg Cu per g (dry wt)
was also detrimental to growth and reproduction (Table 19.2).

Effects of Metals Via the Diet Are More Severe
Than Effects from Chronic Dissolved Exposures
Traditional toxicity tests do not account for the possibility that fish are exposed to metal contamination
via dietary intake. As noted previously, invertebrates in the Clark Fork River have 2 to 100 times greater
cadmium, copper, and lead levels than those collected from tributaries and are an important source of
food for trout in the Clark Fork. Woodward et al. (1994) fed early-life-stage trout invertebrates collected
from near the Clark Fork headwaters and from 85 km downstream. After eating the contaminated diet,
the fish showed elevated concentrations of products of lipid peroxidation and histological abnormalities
(effects on hepatocytes, pancreatic tissue, and the mucosal epithelium of the intestine), as well as reduced
growth and survival (Farag et al., 1994; Woodward et al., 1994, 1995).
This experimental approach was reproduced in a series of experiments using different trout species and
with benthos from different contaminated rivers. Similar toxicological responses usually were observed:
reduced survival (when diets were collected from the Coeur d’Alene River in Idaho), decreased growth,
reduced feeding activity, and histopathological abnormalities (Farag et al., 1999; Woodward et al., 1994,
1995). All of the responses were associated with the bioaccumulation of metals in the tissues of the fish.