Page 971 - The Toxicology of Fishes

Case Study: Pulp and Paper Mill Impacts 951

Robinson (1994), Kovacs et al. (1995), and Kovacs and McGraw (1996) reported that egg production
(fecundity) in fathead minnow was the most sensitive indicator of fish reproductive health; however,
subsequent studies by Borton et al. (2003) showed that, out of 12 bioindicators measured, only time to
spawning (delayed by an average of 10 to 14 days after exposure to PME) was correlated with reduced
egg production. Parrott et al. (2003) found changes in growth and secondary sex characteristics in fathead
minnows exposed long term to final effluent from a bleached sulfite/groundwood mill. Males began to
exhibit ovipositor formation in concentrations as low as 3.2% (v/v), and at higher concentrations most
males externally resembled females. With females, masculinization as tubercle and dorsal fin dot for-
mation was also noted at 3.2% (v/v) and above. Changes in secondary sex characteristics were the most
sensitive indicator of exposure in these tests, which were 4 months in duration (Parrott et al., 2003).

Artificial Stream (Mesocosm) Studies
Field-based artificial stream systems have been used in various forms over the past 30 years to study
algae, benthic invertebrates, and fish (Lamberti and Steinmen, 1993). Studies using fish have examined
growth responses, reproductive behavior, competition, predation, energetics, habitat use, and ecotoxi-
cology (Gelwick and Matthews, 1993). Artificial streams have also been used to assess the effects of
PMEs on fish (Borton et al., 1996; Dubé et al., 2002a,b; NCASI, 1983, 1989, 1993). Artificial streams
have been used for PME assessments in Scandinavia (Tana et al., 2003), the United States (Borton et
al., 1996; NCASI, 1993), and Canada (Culp et al., 1996, 2004; Dubé et al., 2002a,b, 2004) (Table 24.3).
The strength of using field-deployed systems lies in their ability to control exposure conditions while
maintaining some environmental realism. Field-based artificial streams control the type, duration, and
concentration of effluent exposure while maintaining ambient conditions of water temperature, water
quality, and photoperiod (Culp and Podemski, 1996; Culp et al., 1996; Dubé et al., 2002a).
In southwestern Finland, significant work has been conducted examining structural and functional
effects of natural factors (e.g., light, nutrients, temperature) and PME on aquatic ecosystems (Lehtinen,
2003). This work began in 1991 using multispecies brackish water littoral mesocosms consisting of
3
open-air pools (8 m ) supplied with a continuous flow of brackish water from the Baltic Sea (Lehtinen,
2003; Tana et al., 1994).
The National Council for Air and Stream Improvement (NCASI) in the United States has also used
outdoor experimental streams to assess the effects of PMEs (Borton et al., 1996; NCASI, 1983, 1989,
1993). These streams have been in use since 1968 to assess various aspects of PME exposure on different
aquatic species. The streams consist of 100-m in-ground channels that have been modified to represent
different fish habitat types (pools and riffles) and water depths. Some streams are allocated as controls
and others are exposed to effluents. Studies are typically long-term involving exposure durations of
close to a year.
In Canada, various systems have been developed since 1991 to measure PME effects on benthic food
webs (Culp et al., 1996, 2000; Dubé and Culp, 1996). In the late 1990s, the technology developed to
assess the effects of PME on small-bodied fish (Dubé et al., 2002a,b). To provide an example, one
system consists of 16 circular tanks or streams on two trailers (Figure 24.4) (Culp and Podemski, 1996;
Culp et al., 1996; Dubé and MacLatchy, 2000a). The tanks are 1 m in diameter and hold a volume of
227 L. The trailers are situated near a surface water source that is not exposed to PME (i.e., reference
site). Natural receiving water is pumped into a head tank and then through a distribution manifold and
is then delivered to each tank at a controlled rate. PME is also delivered to each tank at a specified
dilution. Each stream is then seeded with substrate and benthic invertebrates and/or fish endemic to the
receiving environment being studied. Experiments are conducted over a desired exposure period (typi-
cally 30 to 60 days) and biological endpoints are measured. In Canada, the most recent phase of
development has occurred since 1999 where multiple systems have been used simultaneously to assess
the effects of PME on benthos, fish, and self-sustaining aquatic food webs (Dubé, 2004) (Figure 24.4).
These advances in artificial stream technology have provided information on effects and stressor sources
that facilitated interpretation of biotic responses measured in the field (Culp et al., 2000). These findings
also contributed to the acceptance of this technology as monitoring alternatives for the Canadian EEM
Programs (Dubé et al., 2002a). Disadvantages of artificial stream techniques include divergence of