Page 764 - The Toxicology of Fishes

744 The Toxicology of Fishes

due to a variety of life-history adaptations, other fish might experience differential exposure to chemicals.
For example, surface-dwelling fish, such as topminnows, would potentially be exposed to high initial
pesticide concentrations found in the surface layer following treatment. Alternatively, contaminants that
sorb to sediments (including many pesticides) might be expected to impact bottom-feeding fish selectively.
Drenner et al. (1993) studied the effects of a pyrethroid insecticide on gizzard shad (Dorosoma cepedianum)
in outdoor microcosms. These fish are filter feeders and commonly have large amounts of bottom sediments
and detritus in their digestive systems. This study (Drenner et al., 1993) was unique in its use of nonstandard
fish species. Similar field studies utilizing other fish species should be pursued to evaluate the influence
of feeding behavior and habitat selection on chemical exposure. Following appropriate research, it is
conceivable that a multispecies assemblage (i.e., surface feeder, water-column planktivore, and bottom
feeder) might eventually be used to better represent potential impacts to natural fish communities.

Chemical Fate Considerations
It must be acknowledged that the primary use of mesocosm testing for fish is to predict the environmental
fate of the chemical. This includes its persistence, its distribution or partitioning among various environ-
mental compartments, and an estimation of its bioavailability and potential to bioaccumulate. (Boyle,
1985). Various chemical characteristics affecting fate are currently measured in the laboratory, such as
solubility, octanol–water and soil–water partitioning, and bioaccumulation in different organisms. More
comprehensive estimates of the fate of the chemical are manifested in mathematical and physical models

of aquatic ecosystems. Boyle (1985) provided a list of different representative types of mathematical
models from the literature used to determine the fate of a potential contaminant. Rand et al. (2000)
described study design, specific techniques, and the fate of pyridaben (a miticide/insecticide) in micro-
cosms and discussed the usefulness of microcosms to study the fate of a chemical under environmental
conditions that are more representative of the field. If the direct effects on growth or reproduction are
desired, a laboratory or early-life-stage test can cost effectively provide that information. The use of
mesocosms is primarily to show that degradation and the metabolic process limit the exposure and, hence,
the availability of the chemical in the environment. Complexity of dosing methods for mesocosm studies
varies with purpose of the study. The contaminant may be added to the water surface, to the subsurface,
or on the sediments by pouring the active ingredient or adding a mixture of soil and toxicant (Boyle et

al., 1996; Cushman and Goyer, 1984; Giddings et al., 1997; Oviatt et al., 1987),spraying with hand-held
sprayers and spanners that release the solution onto the water surface (Brazner and Kline, 1990; Crossland,
1982; deNoyelles et al., 1989; Kedwards et al., 1999a; Stout and Cooper, 1983; Sugiura et al., 1984;
Ronday et al., 1998), or pumping via a flow-through system (Bakke et al., 1988; Farke et al., 1985;
Zischke et al., 1985). Subsurface dosing can also be achieved by placing the spray nozzle or hand-held
sprayer below the water level (Boyle et al., 1985). There are nearly as many application methods as there
are researchers designing microcosm and mesocosm studies. It should be noted that the method chosen
to apply the test material could have considerable influence on its fate and subsequent exposure to
organisms; for example, the droplet size from a spray nozzle of an experimental system may differ from
the droplet size deposited on a natural body of water following agricultural application to adjacent land.
In turn, droplet size may be critical because volatilization from the water surface microlayer can be very
rapid and may be a major route of dissipation (Maguire et al., 1989). Thus, the decision of whether a
chemical is sprayed on the water surface or injected underneath can have a major influence on its half-
life. Clearly, the method of test material application must be chosen so realistic exposures are obtained.

Experimental Design and Statistical Considerations
Experimental Design Considerations
Key issues in designing microcosm and mesocosms tests are treatment replication, sample size and
power, optimization criteria in design selection, choice of number and spacing of dose levels, inference
on safe dose, and defining the dose–response curve (Smith and Mercante, 1989). Biological variables