Page 153 - The Toxicology of Fishes

Toxicokinetics in Fishes 133

C f =  (  kC w) +( k uf Σ PC i)  (3.131)
uw
i
k ex + k eg + k g
where dietary intake is the sum of the proportion of each prey trophic group selected (P ) multiplied by
i
the contaminant concentration in the dietary item (C). The uptake rate from water (k ) is dependent
uw
i
on the initial transfer of a chemical from water to the aqueous phase of the organism and then to the
lipid phase of the organism:
k uw = 1 (3.132)
 (  WQ w) +( W K Q l) 
ow
v
v
where W is the weight of the organism, and Q and Q are transport rates in the aqueous and lipid
w
v
l
phases. Q is estimated from an empirical relationship that relates chemical transport to body weight:
w
Q w = 88 3 W v 06 . (3.133)
.
Q is less well known than Q but is thought to be about 100 times smaller than Q :
w
w
l
Q 1 = Q w 100 (3.134)
The uptake rate from food (k ) is dependent on dietary absorption efficiency (α, the fraction of the
uf
dietary dose that is absorbed by the animal), feeding rate (FR), and W :
v
k uf =α FR W v (3.135)
The FR depends on both allometric and bioenergetic considerations and may be estimated from an
empirical relationship based on body weight and ambient temperature (T in °C; annual average):

.
.
e
.
FR = 0 022 W v 085 006 T (3.136)
The excretion rate of a chemical across the gills (k ) is dependent on the transfer of chemical from lipid
ex
to the aqueous phase of the fish and then to ambient water:
k ex = 1 (3.137)
 (  WL K ow Q w) +( WL Q l) 
v
v
i
i
The fecal egestion rate (k ) is calculated as 0.25 times the uptake rate from food:
eg
.
k eg = 025 k uf (3.138)
The growth of a fish over time tends to reduce contaminant concentrations by increasing the tissue mass
into which chemicals are distributed. This phenomenon is referred to a growth dilution and is accounted
for by calculating k as a weight-dependent coefficient:
g
.
k g = 0 000502 W v −02 . (3.139)
Food web models are typically calibrated using existing data on contaminants in an ecological system
similar to that which is being modeled. Parameter inputs, including body size, temperature, prey selection,
lipid content, and asorption efficiency, may be determined from literature values or site-specific mea-
surements. As with compartmental and PBTK modeling, parameter values can be adjusted to obtain
better agreement between observed and model-estimated concentrations of contaminants in biota. As