Page 974 - The Toxicology of Fishes

954 The Toxicology of Fishes

for example, coordinated a survey of fish responses at multiple mills and concluded that responses of
MFO induction, reduced gonad sizes, and altered steroid hormone levels could be seen at mills with a
variety of process types and at some sites with secondary effluent treatment. Additional studies examined
the relationship between MFO induction and effluents from kraft mills with different bleaching and
effluent treatment technologies (Martel and Kovacs, 1997; Martel et al., 1996, 1997; Williams et al.,
1996). Results showed that MFO-inducing potential could not be linked to a particular bleaching or
pulping process. Induction was present in unbleached kraft mills, bleached kraft mills with different
bleaching sequences, thermomechanical mills, and chemithermomechanical mills. In addition, induction
was not correlated to typical compound classes including resin and fatty acids, dissolved organic carbon,
adsorbable organic halogen (AOX), or other chlorinated organics. These results suggested that the source
of the responsible compounds was independent of the bleach plant on the pulping side of mill operations
(Folke, 1996; Kovacs and Megraw, 1996; Munkittrick et al., 1998). Williams et al. (1996) speculated
that the origin of the inducing compounds was residual lignin liberated from the cellulose in either the
bleaching process (through oxidation) or in the pulping process and that the discharge of pulping liquors
(i.e., weak black liquor) could be a source of inducers in mills.
Examination of specific mill process streams by Martel et al. (1997) identified contaminated conden-
sates produced when weak black liquor is reduced through evaporation in the chemical recovery process,
as a major process source of MFO-inducing substances. Hodson et al. (1997) and Schnell et al. (2000)
identified weak black liquor from softwood pulping as a primary source of inducing compounds. Artificial
stream studies conducted by Dubé and MacLatchy (2000a) before and after a process change (reverse-
osmosis treatment of condensates) showed that sex steroid depressions were removed in an estuarine
killifish (Fundulus heteroclitus) after the process change. This result suggested that condensates were
also a source of steroid-disrupting compounds. Acute and sublethal toxicity testing (Dubé and MacLatchy,
2000b) and laboratory exposures (Dubé and MacLatchy, 2001) using these mill waste streams confirmed
that condensates were a source of steroid-disrupting compounds, and reverse-osmosis treatment removed
the effect (Dubé and MacLatchy, 2001).
A similar waste stream assessment approach was used by Parrott et al. (2000c) in an attempt to isolate
the origin of reproductive effects from effluents discharged from a bleached sulfite mill and a bleached
kraft mill. Exposure of goldfish for 21 days showed that final effluents at 100% caused sex steroid
disruptions, but the source of the effects could not be identified. Subsequent goldfish exposures after a
process change at the bleached sulfite mill showed that the steroid-disrupting effect was removed (Parrott
et al., 1999). The nature of the process change was not reported. Studies on the Miramichi River in New
Brunswick, Canada, also showed endocrine disruption activity in mummichog (Fundulus heteroclitus)
exposed to an environmentally relevant concentration of secondary treated final effluent. Sex steroids
concentrations were reduced in these fish to a level greater than that observed with exposure to primary
treated effluent at the same concentration, suggesting that secondary treatment itself was a possible
source (Dubé et al., 2002b). Although this waste stream approach has been effective at isolating sources
of compounds causing MFO induction and sex steroid depressions in fish, results are very mill specific.
Each pulp mill is unique in its handling of waste streams; thus, a detailed understanding of mill operations
and mill-to-mill differences in operations is required before the relevance of these results among mills
can be determined. Further studies are underway in an effort to identify the source of the reproductive
effects at these mills.

Characterization and Identification of Causative Compounds
The last tiers of the Hewitt et al. (2003a; 2005a) framework relate to characterizing the chemical classes
involved in the effect and, ultimately, identifying the specific chemicals associated with the responses.
The exposure profile of the effect can indicate several properties of the chemicals; for example, if the
response in aquatic biota occurs rapidly upon exposure, it indicates the responsible compounds are
readily bioavailable. To illustrate this concept, we present a caging study by Munkittrick et al. (1999)
implemented to understand the characteristics of the PME compounds responsible for MFO induction.
MFO activity was induced after 3 to 4 days. In addition, (1) uptake of the inducer was very rapid in