values were set to equal values; no ratio was calculated for ND values. The 
results show that the mean concentrations of Ag, Pb, and Zn in fried fish 
were substantially higher than the corresponding values in raw fish. The 
ratios of FF/RF were 10.1 +/- 0.86, 1.8 +/- 1.8, 1.1 +/- 0.95, and 1.5 +/- 
0.8, respectively. The FF/RF ratio for As was 0.8 +/- 0.3 which shows a 
"slight" decrease in the concentration of As as a result of frying. The 
lower values of some metals, such as As and Hg, in fried fish may have 
been due to the presence of volatile metal compounds (methylated forms of 
As and Hg) that were lost from the tissue during frying of the samples. 
Higher values of some metals, such as Ag and Pb, in fried fish should be 
interpreted as indicating no substantial change in concentration. 
Level 1 trace organics analysis . Spiked recovery of hexachlorobutadiene 
in eight species averaged 140 +/- 35%. This result along with the 
higher variability seen in replicate samples, suggests that interference 
may have been significant in low level samples. The precision of instrumental 
2 
analysis was 1.9% RSD across the calibrated range, with an R value of 
0.9999 (for quadratic response function). 
Spiked recovery of hexachlorobenzene in eight species averaged 106 +/- 18%. 
In the five replicate raw fish analyses where HCB was detected, good 
agreement between analyses was seen in three instances, while in two cases 
(0.7 and 0.8 ppb), the replicate level was below the detection limit. In 
actual fish samples, HCB was found above the detection limit in 21 of 67 
samples, with a range and average concentration of 0.5-8.0, and 1.5 ppb, 
respectively (Table 3). The levels seen are in general agreement with 
previous results. 
Spiked recovery of p ,p'-DDE in eight species averaged 93.4 +/- 18%. 
Replicate analysis of seven fish having detectable DDE showed good agreement 
in five cases, with two examples having less than detection limit results 
in one replicate. A closely related compound, o,p'-DDE was used as an 
intra-assay recovery standard, and showed average overall recovery of 
80.6%, with a standard error of 1.8%. DDE was detected in 59 of 67 fish 
samples (Table 3), with a range and average amount of 0,93-15.6, and 3.6 ppb, 
respectively. This range of values corresponds reasonably well with that 
of previous studies. 
The compounds o,p'-DDD and o,p'-DDT are not expected to occur to any 
significant extent in environmental samples, as the commercial DDT used 
and introduced as pollution was largely the p,p' isomer. The o,p'-DDT 
isomer co-elutes with the p,p' isomer of DDT under the GC conditions used 
in this study, so these agents are reported together. However, it is 
reasonable to infer that all of the detected pesticide is contributed from 
the p,p'-DDT. The quality control results for the o,p'-DDD were detected 
in 10 of 67 samples, with a range and average amount of 0.75-5.7, and 
1.8 ppb, respectively. None of these low level "hits" were confirmed in 
the GC/MS aiialysis. Given the method detection limit of approximately 
0.7-1.0 ppb for o,p'-DDD, the few examples of its detection in these 
samples were probably analytical artifacts. 
123 
