Page 966 - The Toxicology of Fishes

946 The Toxicology of Fishes

• Statistical design criteria including hypotheses, statistical methods, determination of sample
size (statistical significance and power analysis)
• Description of how other supporting information (e.g., changes in benthic invertebrate com-
munities, effluent chemistry data, toxicity data) will be used in the assessment of fish effects
• Identification of the quality assurance and quality control measures that will be taken to ensure
the validity of the data

One of the most difficult decisions to make before a study is conducted is to define what an effect will
be and how it will be measured. It requires identification of indicators to measure at some level of
biological organization and an experimental design where changes in the indicators can be assessed
relative to some benchmark or baseline condition. It also requires a framework to decide what information
will act as supporting data and how they will be used to assist with interpretation of effects on fish. A
good starting point for information and references on indicator selection and study design trade-offs can
be found in Munkittrick et al. (2000b) and Environment Canada (1998, 2001).
Pulp and paper effects assessment with fish in Finland, Sweden, New Zealand, and Canada has made
a significant contribution to development of study design approaches; for example, in the Canadian EEM
program, mills are required to conduct biological monitoring studies that include a fish population and
a benthic invertebrate community survey. They are also required to collect supporting information through
effluent and water quality monitoring studies, which include effluent characterization, water quality
monitoring, and sublethal toxicity testing (Environment Canada, 1998). The fish population survey is
conducted to determine if there have been changes in indicators of fish growth, reproduction, condition,
and survival by collecting fish species found in exposure and reference areas and comparing measure-
ments of length, weight, gonad size, liver size, fecundity, and egg size. In this program, an effect on the
fish population means a statistical difference between fish population measurements taken from exposure
and reference areas (control/impact study design). Fish studies have been conducted at 65 mills across
Canada using core effect indicators and statistically rigorous study designs, resulting in a national
synthesis of effects (Environment Canada, 2003; Lowell et al., 2004). The database from this investigation
has led to recognition of key common response patterns in fish and has presented a context to begin
development of critical effect sizes for each indicator.

Relevance of Other Trophic Levels
Although this chapter focuses on fish responses to PMEs as a case study, it is important to consider
other trophic levels when designing a fish assessment program. Aquatic food webs include organisms
at a wide variety of organizational levels, and PME assessments have been conducted focusing on lower
trophic levels (e.g., algae and benthic invertebrates), as well as higher levels (e.g., small- and large-
bodied fish) (Environment Canada, 1998; Munkittrick and McCarty, 1995; Munkittrick et al., 2000b).
All organisms integrate the stressors within the system, and all levels of organization are capable of
exhibiting measurable changes in their own right; however, the monitoring level selected must be relevant
to the hypothesis to be tested and/or to the decision-making process used to manage ecosystem health.
Many of the large international field programs conducted to date have been designed as multi-trophic,
integrated studies that examine PME effects on algae, benthic invertebrates, and fish (Andersson et al.,
1988; Larsson et al., 2000; Lowell et al., 2004; Owens, 1991; Södergren, 1989). This approach provides
invaluable information on different compartments of the aquatic receiver and can build a weight-of-
evidence to determine if PME effects exist, if the effects are common across trophic levels, and to assist
with interpretation of the mechanism of the effects (Owens, 1991). For example, in the national synthesis
of the Cycle 2 of the Canadian EEM program, a response pattern observed in biota exposed to PME
included increased benthic invertebrate density and species richness, as well as increased energy storage
(i.e., increased condition and liver size) and energy allocation (i.e., increased gonad size and growth) in
fish (Lowell et al., 2004). Often these responses were also coincident with increased algal biomass.
These results strongly suggest a nutrient enrichment effect of the effluent across trophic levels. Once a
system is understood, then monitoring multiple trophic levels may be neither necessary nor cost effective