Significantly higher concentrations of Aroclor 1254, Aroclor 1260, and a sum of three 

 aroclors were found in the OK and BK fish. Aroclor 1242, in contrast, was higher in RR fish 

 than in BK fish. About one-third of the PCB congeners (IUPAC 8, 44, 66, 153, 101, 206) 

 accounted for 58 percent of the variance among the PCBs. 



Lipid-normalized concentrations of contaminants in liver were generally not related to 

 length or weight of the fish. Also, no significant correlations were apparent between gonad 

 weight, liver weight, GSI and HSI and the concentrations of contaminants. These 

 observations pertained to both the complete collection of November-December fish and to 

 fish collected at individual sites. 



No consistent pattern in contamination was apparent for all the quantified chemicals 

 among the sites. Fish from one site with the highest concentrations of one group of 

 compounds did not necessarily have the highest concentrations of the others. For example, 

 fish from the BK and OK sites had the highest tPCB concentrations, but among the lowest 

 concentrations of heptachlor epoxide. Fish from the VJ site had relatively high DDE 

 concentrations, but concentrations of other pesticides in those fish were relatively low. Fish 

 from the RR site were surprisingly contaminated with PCBs relative to those from VJ and 

 SP. However, the RR fish were larger than those from the other sites. The relative 

 similarity in mean contaminant levels between BK and OK reflected the proximity of these 

 two sites to each other. Overall, the fish from the SP, RR, and SC sites were generally the 

 least contaminated, and those from the BK and OK sites were generally the most highly 

 contaminated. Within-site variability was relatively high, as indicated by high standard 

 deviations relative to the means. Some degree of variability in contaminant levels is to be 

 expected in a population of feral fish because of their mobility. However, it is possible 

 that these fish were exposed to and were affected by readily metabolized and nonquantified 

 compounds, such as aromatic hydrocarbons, that may have been partially or wholly 

 responsible for the induction of the biological measures. If the latter case pertained, then a 

 strong correspondence between the biological data and the quantified chemical data would 

 not be expected. 



72 



