Page 814 - The Toxicology of Fishes

794 The Toxicology of Fishes

7500

Metal Concentration (nM/g) 2500

500

75

25

5
Hydropsyche Baetis Serratella
Arctopsyche Epeorus
FIGURE 19.10 Difference in copper (triangles) and cadmium (squares) bioaccumulation in the five species followed in
Figure 19.9. The species that progressively disappear with increased contamination also bioaccumulate 5 to 10× more metal
than the species that appear, by their ecological distributions, to be more tolerant to the metals (the y-axis scale is log-
converted to allow presentation of both metals).

Observations of ecological change, alone, are inadequate to convincingly demonstrate that metals are
causing adverse effects on benthic communities, especially where a great deal is at stake, as at this
Superfund site. Other stressors in the watershed (e.g., dewatering, temperature, nutrient inputs) might
also correspond with the metal gradient. Mechanistic explanations can improve the explanatory power
of a correlation. It is generally accepted that species that bioaccumulate or detoxify metals differently
ultimately have different sensitivities to metals (Rainbow, 2004). Cain et al. (2004) showed that bioac-
cumulation was reduced and detoxification capabilities were enhanced in Hydropsyche species and
Arctopsyche grandis (species deemed tolerant from their distribution in the river) (McGuire, 1999).
Bioaccumulation was considerably greater in species typically absent in contaminated waters, including
such mayflies as Serratella tibialis and Timpanoga species (Figure 19.10). Some sensitive species (e.g.,
S. tibialis) do not sequester metals as efficiently into detoxified fractions within their cells (metal-specific
binding proteins or intracellular granules) as do the Hydropsyche (Cain et al., 2004). Those species
(especially mayfly species) that tend to accumulate high concentrations of internal copper and cadmium
in forms available for binding to sites active in causing toxicity also tend to be absent from contaminated
sites. These findings specifically linked species absences with the likelihood of metal effects.
In contaminated areas, macro-invertebrate drift can increase, community and microbial respiration
can decline, and leaf litter breakdown declines (Carlisle and Clements, 2005). Carlisle and Clements
(2003) concluded that total production attributable to algae and animal prey declined in contaminated
streams. Impairment of moderately sensitive species can include effects as subtle as changes in predator
avoidance (Lefcort et al., 2000). None of these effects has been studied in the Clark Fork; nevertheless,
such results suggest that fish inhabiting a metal-contaminated stream must depend on a disturbed benthic
community for their food—one that is a contaminated food source, less productive, less diverse, and
missing attributes of likely (but usually unquantified) importance to fish diversity and productivity.

Effects on Fish
As noted above, the traditional approach to evaluating risk to fish is a comparison of ambient dissolved
concentrations against the concentration of metal that elicits effects in laboratory tests. Verification of
effects from conditions in the water body are not required for regulatory compliance but are ultimately
desirable to justify policy judgments. Interpreting effects on fish in the field is complex, however, and
requires a suite of repeated observations building to a systematic body of evidence pointing toward cause
and effect. Useful lines of evidence include documented fish kills, in situ toxicity testing, population