between Cd?+ exposure and no Cd” exposure (Cd?> 
treated fish were matched by weight +30% with non- 
treated fish within each experiment). Wilcoxon’s two- 
sample test for unpaired observations was used to test 
significance between 30- and 90-min time intervals. 
Figure 1 shows that within the first 30 min, Cd** ex- 
posed fish clear bacteria from the blood stream more 
rapidly and take up greater numbers of bacteria in the 
liver and spleen than fish not exposed to Cd** 
Differences in counts in the blood and liver between 
Cd?+ treated and nontreated fish are significant. 
However, counts in the cadmium-treated fish change 
very little during the next 60 min, i.e., only 6.3% ad- 
ditional cells are eliminated. By contrast, the counts 
in fish not exposed to cadmium continue to decrease 
in the blood, liver, and spleen between 30 and 90 min. 
Values for the blood and the total remaining cells in 
the blood + liver + spleen are significantly different; 
28.7% of the initial bacterial load is eliminated (or 
killed) within the 60-min time interval. Thus, at 90 
min the total remaining bacteria in the non-Cd?*fish 
are significantly lower (P.01) than in the Cd** treated 
fish. Statistical analyses were done on all possible 
relationships between elements of the 0 ppm and 12 
ppm Cd**treated fish and between the 30- and 90-min 
time intervals. Any probabilities not shown were 
found to be nonsignificant. 
DISCUSSION 
From the experimental data presented in this 
paper, one can conclude that short-term exposure of 
the cunner to CdCl, at toxic or near toxic levels does 
not affect the production of antibody against SRBC. 
Although this conclusion is based upon early antibody 
production which, undoubtedly, had not reached a 
peak, it seems safe to assume that significant 
differences, which were to appear, would show up as a 
delay or lag in these early responses. Fish could not be 
held for long periods of time because of limitations in 
holding space. Although the second SRBC injection 
(to hasten the rise in antibody titer) was given after 
fish had been in Cd?-free water for 3 days, it is certain 
that fish still had high Cd?* levels at this point. The 
data of Greig, Adams, and Nelson (this report, Part 
Il) show continued high levels of Cd?*+in cunners after 
4 wk of holding in Cd?+-free water. Others have shown 
that the half-life of Cd** after a single exposure does 
is in excess of 200 days in rats, mice, dogs, and mon- 
keys (Friberg, Piscator, and Norberg, 1971, p. 66). 
The low-level agglutination titers observed in about 
half the nonimmunized fish (see titers in Table 1) are 
not unusual. Natural or nonspecific agglutinins are 
common among fish, as well as other animals. This 
does not interfere with immunization experiments as 
long as these agglutinins are low enough in titer that 
they are not confused with the results of specific im- 
mune stimulation. 
In contrast to the results on antibody production, 
Cd** did have significant effects on uptake and 
destruction of bacteria by phagocytes in the liver and 
spleen. Cadmium at 12 ppm was used in these studies 
because it was the highest level at which there was 
consistent survival of fish during post-Cd?*+ holding. 
Fish exposed at this level exhibited a more rapid in- 
itial uptake of bacteria by cells of the liver and spleen, 
but a slower bacterial destruction rate than fish not 
exposed to Cd** These results are consistent with con- 
clusions drawn by Holmes, Page, and Good (1967) 
that the metabolic events accompanying phagocytosis 
can be separated into two categories: 1) events 
associated with particle uptake and 2) events 
associated with degranulation within the phagocyte. 
In the present study it appears that Cd?* stimulates 
the metabolic events responsible for bacterial uptake 
but inhibits degranulation or those events responsible 
for delivering bactericidal substances to the inter- 
nalized bacteria. 
The initial, relatively rapid clearance of bacteria 
from the blood cannot be entirely credited to 
phagocytosis in the liver and spleen since phagocytic 
cells in the kidney and gill tissues could also con- 
tribute to bacterial uptake and destruction. However, 
cells of the liver and spleen probably take up the ma- 
jor portion of the injected antigen. This is assumed for 
two reasons: 1) in some instances, the liver and spleen 
contained as much as 80% of the tota! bacteria initial- 
ly injected and 2) studies in other animals indicate 
that cells of the liver and spleen are responsible for 
removing the majority of intravenously injected par- 
ticulate antigens. Benacerraf, et al. (1957) found that 
the liver and spleen of rats removed 85 to 98% of in- 
jected carbon or saccharated iron oxide. McCloskey 
(1972) showed that the liver and spleen of mice retain- 
ed higher proportions of injected bacteria than other 
organs. However, data in the latter two references 
show the liver as the organ of major uptake; whereas 
in cunners, the spleen usually contains more bacteria 
than the liver. Since the anterior kidney has been 
shown to be a site of antibody production in rainbow 
trout (Chiller et al., 1969), it is assumed that this 
organ may also take up significant numbers of 
bacteria; however, for the reasons already given, it is 
unlikely that this uptake in the cunner of is the same 
magnitude as that of the liver and spleen. 
Lack of significant differences between levels of 
bacteria in the liver and spleen at 30 and 90 min (as 
seen in Fig. 1) does not indicate lack of activity within 
their phagocytic cells. As bacteria are destroyed 
within loaded phagocytes, additional bacteria can be 
taken from the blood stream to reload the phagocytes. 
Hence, the bacterial levels in these organs may appear 
to be static when, in fact, there is a rapid turnover. 
The destruction rate of the bacteria (28.7%/hr in the 
normal fish) can be greatly increased when bacteria 
from 18-hr growth cultures are used (rather than 72-hr 
