Page 876 - The Toxicology of Fishes

856 The Toxicology of Fishes

Probability
What is the probability of HAHs causing such an effect in lake trout of the Great Lakes? This criterion
is viewed not necessarily in the statistical sense but in the biological sense of probability. Could it be
probable that HAHs elicit such an effect in lake trout? In this sense, effects of HAHs on lake trout
populations were most certainly a possibility and indeed a probability. Lake trout were exposed to elevated
concentrations of HAHs from the initial production and release of these chemicals into the Great Lakes
ecosystem until after they were banned in the late 1970s (Figure 21.7 and Figure 21.8). Lake trout are
among the most sensitive fish species tested with HAHs (Elonen et al., 1998; Spitsbergen et al., 1991;
Walker et al., 1991), with embryo and fry mortality occurring at doses as low as 40 pg TCDD per g egg
(Spitsbergen et al., 1991; Walker et al., 1991, 1994) and sublethal effects of hemorrhage and yolk sac
edema occurring at doses as low as 2 to 15 pg TEQ per g egg with environmental mixtures of HAHs
(Tillitt and Wright, 1997; Wright, 2006). Both measured (DeVault et al., 1986, 1989) and estimated
(Cook et al., 2003) concentrations of TEQs in the eggs of Great Lakes lake trout exceeded these thresholds
for toxicity during the three-decade period in question. Thus, strong direct evidence and plausible logic
indicate a reasonable probability for HAHs/TEQs to have had adverse effects on lake trout reproduction
and development during this period; therefore, the criterion of probability appears to be met.

Time-Order
Did the production and release of HAHs into the Great Lakes environment precede the adverse impacts
observed on lake trout reproduction? Production of PCBs was first reported in Anniston, Alabama, and
the first large-scale production of PCBs in the United States was by the Monsanto Corporation at their
Sauget, Illinois, facility in 1929 (Durham and Oliver, 1983). PCBs were used initially by industry as
dielectric fluids in capacitors and electrical transformers, then their popularity increased and their use
®
spread to other applications. The annual usage rate of the most popular PCBs (Arochlor 1242) increased
in the United States to 18 million pounds/year in 1957 and over 50 million pounds/year by 1970 (Cairns
et al., 1986). The presence of PCBs, PCDDs, and PCDFs in the Great Lakes has been documented in
the sediment record since the 1930s (Cook et al., 2003; Durham and Oliver, 1983). The most precise
and presumably accurate account of HAH concentrations and exposure in Great Lakes sediments during
the 20th century was described for Lake Ontario (Cook et al., 2003). Concentration estimates of HAHs
and TEQs in lake trout from approximately 1930 to 1990 indicated that toxicity thresholds of TEQs in
lake trout were exceeded as early as 1940. By 1950, in Lake Ontario the concentrations of TEQs in the
eggs of lake trout were great enough to cause 100% mortality in the resultant embryos or fry (Cook et
al., 2003). This timeframe coincides with the extinction of the lake trout from the lower Great Lakes,
the early 1950s in Lake Ontario (Figure 21.9) and Michigan (Figure 21.2).
Of course, heavy commercial fishing pressure depleted stocks of adult lake trout in all of the Great
Lakes, including Lake Superior, but only in the lower Great Lakes of Ontario, Michigan, and Huron
were lake trout extirpated. Lake trout in Lake Superior always maintained self-sustaining populations
throughout this period. The predation of adult lake trout by sea lamprey in Lake Superior was less due
to smaller populations of the lamprey in Lake Superior (Smith and Tibbles, 1980). Lower rates of sea
lamprey predation in Lake Superior may have been an additional factor in the survival of those popu-
lations; yet, only in the lower Great Lakes, where the concentrations of chemical contaminants exceeded
thresholds for toxicity to lake trout, did the populations of lake trout fail completely. Even stocking of
hundreds of thousands of lake trout into the lower Great Lakes during the 1970s was not sufficient to
bring about a self-sustaining population. So, based on the temporal coincidence of rises in dioxin-like
contaminant concentration and increases in reproductive failure of lake trout in the lower Great Lakes,
the time-order criterion is supported.
The criterion of time-order is also supported by the biological data on monitoring of hatching and fry
survival of lake trout from Lake Michigan and Lake Ontario. Mac and Edsall (1991) summarized the
temporal trends of hatching and fry survival over the period from 1975 to 1988. Lake trout eggs collected
from adult fish from southeastern Lake Michigan had a clear increasing trend in survival from 1975 to
1988, roughly 65% to 100% over this period (Mac and Edsall, 1991). Fry survival over this same period
also had a similar increase, with the exception of 1975 and 1977; however, protocols in the early years