Page 643 - The Toxicology of Fishes

Toxicity Resistance 623

differing in amino acid sequence. Allozymes used in population genetic studies are generally those
involved in biochemical respiration. Variation observed in stressed populations is therefore not presumed
to be mechanistically linked to survival, but may co-occur with genes linked to survival. Nucleic acid
techniques involve amplification and detection of random, repeated, or known gene fragments and
comparing their occurrence in different fish populations (Bagley et al., 2001; Sunnucks, 2000).
Numerous field studies support the hypothesis that environmental contaminants can alter population
structure and genetic diversity in fish. Reduced genetic diversity has been demonstrated in fish popula-
tions associated with a uranium processing facility in Ohio, a coal-ash settling basin in South Carolina,
and an acid-impacted drainage in New York (Guttman, 1994). Murdoch and Hebert (1994) reported
reduced mitochondrial DNA variation in brown bullhead from contaminated sites relative to those from
less contaminated sites in the Great Lakes. Gillespie and Guttman (1989) found significant differences
in allele frequencies in populations of the central stoneroller (Campostoma anomalum) from sites above
and below a contaminated region of a stream.
Kopp et al. (1992) analyzed genotypic frequencies in central mudminnow (Umbra limi) populations
from acid-stressed and nonacid-stressed sites and found that for several loci the stressed populations
were characterized by a much greater frequency of one particular allele. The results suggested that
environmental conditions were acting as selective forces; however, these data must be interpreted
cautiously, because population differentiation can result from processes other than pollutants acting as
selection agents (Diamond et al., 1991; Gillespie and Guttman, 1989; Kopp et al., 1992; Mulvey and
Diamond, 1991). Studies of 20 polymorphic loci in mummichog, for example, revealed significant
directional changes in gene frequencies with latitude (Cashon et al., 1981: Powers and Place, 1978;
Powers et al., 1986; Ropson et al., 1990). Thus, although different allozyme patterns were found in
populations of mummichog from polluted and nonpolluted sites in a study by Heber (1981), these
differences were associated with the north–south pattern in genotypes of this species and apparently
were not pollution related.
The hypothesis that changes in the genetic structure of a population can be attributed to environmental
contamination is supported by results of several laboratory exposure studies. Gillespie and Guttman
(1989) found that stonerollers with genotypes that appeared to be sensitive to contaminants in the field
also appeared to be more sensitive to copper toxicity in laboratory tests. Chagnon and Guttman (1989)
exposed laboratory stocks of mosquitofish to copper and cadmium and found survival to vary significantly
with different allozyme genotype and specific alleles. Diamond et al. (1989) and Newman et al. (1989)
examined genotypic frequencies at eight enzyme loci during acute exposure of mosquitofish to arsenate
or inorganic mercury and found that time to death was related significantly to genotypes at two of the
eight loci for arsenate and three of the eight loci for inorganic mercury. Other investigators reported
differential survivorship among mosquitofish genotypes following exposure to mercury (Diamond et al.,
1989), arsenate (Newman et al., 1989), and cadmium (Chapman and Guttman, 1989). Differences in
time to death were also reported in different genotypes of mosquitofish and sand shiner (Notropis
ludibundus) exposed to the pesticides lindane and parathion (Sullivan and Lydy, 1999).
Mummichog in the Elizabeth River in Virginia and New Bedford Harbor in Massachusetts have
provided opportunities to evaluate the effects of contaminant exposure on genetic diversity and population
structure in populations of DLC- and creosote-resistant mummichog. Allozyme analyses were conducted
in mummichog collected from sites in the Elizabeth River and York River that vary by several orders of
magnitude in concentrations of PAHs (Mulvey et al., 2002, 2003). Fish from the most heavily contam-
inated site (Atlantic Wood) were genetically distinct from fish collected from other Elizabeth River sites.
The data suggested that a locally stable population existed at the Atlantic Wood site; however, no evidence
for decreased genetic diversity in this population was observed. Further, no relationship was observed
between sediment PAH concentrations and genetic diversity in fish collected from any of the study sites.
Similar results were reported in studies with New Bedford Harbor mummichog (McMillan et al.,
2006; Roark, 2003). Differences in the genetic structure of the populations under investigation were
related to the geographic distances separating them. No relationships were observed among genetic
structure, sediment contamination at the collection sites, and PCB tolerance in the populations inhabiting
those sites. Genetic diversity did not differ significantly between sites, and the New Bedford Harbor fish
did not exhibit a reduction in genetic diversity relative to those collected from relatively clean sites. In