Page 111 - The Toxicology of Fishes

Toxicokinetics in Fishes 91

intraperitoneal injection or dietary exposure (Varanasi et al., 1978). At later time points, only naphthalene
metabolites were found. Whether or not these metabolites were formed in skin prior to their diffusion
into mucus was not investigated. Using an in vitro skin strip preparation, Ali et al. (1987) showed that
the skin of the African catfish (Clarias gariepinus) exhibits both phase I and phase II metabolic activity
toward some endogenous steroids.

Elimination Via the Gametes
The redistribution of lipophilic chemicals from somatic lipid stores to developing gametes was described
earlier in this chapter. Compounds for which this phenomenon has been demonstrated include anthracene,
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), DDT and its metabolites, PCBs, dibenzofurans, mirex,
heptachlor epoxide, toxaphene, chlordane, and dieldrin (Kleinow et al., 1999). Additional studies have
demonstrated the accumulation of zinc, lead, mercury, and cadmium in gametes of several fish species
(Morrison et al., 1985). Vitellogenin binds several micronutrient metals, including iron, copper, and zinc;
however, little is known about its role in transporting pollutant metals to developing oocytes (Kleinow
et al., 1999).
The triggers for gamete-based elimination are gonadal maturation and reproductive behavior. These
events may occur only seasonally for some fish species and year-around for others. Spawning may
consist of a single event or multiple events (fractional spawners) occurring over a short or extended time
interval. Although spawning eliminates xenobiotics from the animal, reducing the whole-animal con-
taminant load, whole-body chemical concentrations may increase (Tietge et al., 1998), decrease (Guiney
et al., 1979; Niimi, 1983; Vodicnik and Peterson, 1985), or stay the same (Niimi, 1983), depending on
the chemical concentration in gametes relative to the rest of the body. If the contaminant concentration
in the gametes is lower than that of the rest of the fish, gamete release actually increases the chemical
concentration in the remaining tissues (expressed on a whole-body basis). The reverse is true if the
contaminant becomes highly concentrated in the gametes, relative to the rest of the animal.
The importance of maternal transfer in viviparous species is currently unknown. Wourms et al. (1988)
estimated that, among teleosts, live-bearing strategies have developed in only 2 to 3% of species; however,
two of these species, the mosquitofish and guppy, have been used extensively in aquatic toxicology
research. To the extent that an exchange of blood occurs between female fish and the developing young
(as in Poeciliidae), it can be speculated that maternal transfer of chemicals to offspring occurs in some
live-bearing species. In species for which most early development takes place after the egg has left the
follicle (Embiotocidae and live-bearing sharks), the opportunity for maternal transfer may be reduced.

Toxicokinetic Modeling

The uptake and disposition of xenobiotic chemicals in fish have been described quantitatively using both
compartmental and physiologically based modeling approaches. Compartmental models are comprised
of one or more compartments, each of which represents tissues and organs with similar kinetic properties.
These compartments do not generally correspond to any particular anatomical or physiological entity.
Physiologically based models also represent the fish using a set of compartments. The structure of the
model, however, is based on anatomical, physiological, and biochemical information for the species of
interest. The compartments in a physiologically based model often correspond to specific tissues and
organs, although kinetically similar tissues may be lumped to form a single compartment. In a compart-
mental model, parameter values and the number of compartments are determined by nonlinear regression
analysis of measured kinetic data. Physiologically based models may, in theory, be developed in the
absence of experimental data. In practice, however, measured values are often required to fit a small
number of critical parameters.
Both modeling approaches may be used to project toxicokinetic behavior beyond the experimental
dataset on which the model is based. Using compartmental models, this may be done by determination
of model parameter values for fish of different sizes or for different values of an environmental variable
such as temperature. Empirical relationships that relate model parameters to fish size or an environmental