Tukey's test revealed that no between-station differences 

 occurred among the three mussel populations (WLISc, 500MW and 

 WLISrN) and that the Aroclors and PCB concentrations were 

 significantly less in the Rlr population than in the other three. 

 The concentrations of Aroclor 1242 and total PCBs were different 

 before and during disposal in the mussel populations placed at 

 WLISrN and WLISc; both compounds were significantly higher during 

 disposal operations (p < 0.046 to 0.001). The findings of 

 significant temporal and spatial variations in Aroclors and total 

 PCB concentrations are similar to that of the trace metal data 

 (Cd, Cu> Zn, Fe) discussed in the previous section. 



Stepwise multiple regression analyses (Table 7) show that 

 the volume of dredged materials was entered only into the 

 disposal site stations (WLISc and 500MW) . At WLISc, the volume 

 of dredged materials disposed was the first variable entered into 

 the regression model and accounted for 73.6% of the variance in 

 Aroclor 1242. Sixty to 80% of the variance of other Aroclors and 

 total PCBs could be explained by such intrinsic factors as W/D 

 ratio and shell length. Similar observations were made at 500MW, 

 i.e., 50-75% of the variance in Aroclors 1242, 1254, 1254+1260 

 and total PCBs were attributable to the intrinsic variables. 

 However, the dredged volume did account for 9, 13, and 21% of the 

 variance in total PCBs, Aroclor 1260 and Aroclors 1254+1260 

 respectively at this station. Again the analyses suggest that 

 the disposal operation was not a major factor for the uptake of 

 PCBs in the monitoring populations. This observation is 

 consistent with changing levels of PCBs associated with dredging 

 operations reported in San Francisco Bay (Anderlini et al. , 

 1975) , Puget Sound (Engler, 1979) , Eastern Long Island Sound 

 (Arimoto and Feng, 1983a) , and Central and Western Long Island 

 Sound (Feng, 1984, 1985). 



A significant compositional change in tissue Aroclors in 

 favor of higher chlorinated Aroclors (1254+1260) were first 

 reported in 1984 (Feng, 1984) when the mussels were transplanted 

 from LATr to CLIS and WLIS disposal and reference sites. Such 

 changes have again been evident in the present study (Table 8) . 

 The percent relative concentrations of Aroclors 1242 and 

 1254+1260 showed significant station-differences as revealed by 

 one-way ANOVA (F=6.15, d.f.=3,39, p<0.0016). Further analyses of 

 the data by Tukey's test indicate that the Rlr mussel population 

 was significantly different from the other three mussel 

 populations in their relative tissue concentration of Aroclor 

 1242 and Aroclors 1254+1260 (Table 8) . It is apparent that the 

 45:55 Aroclor 1242/Aroclor 1254+1260 ratio for the Rlr population 

 differed markedly from 33:67 for the other three populations. 



Bis (2-ethyl hexyl) Phthalate . During the analysis of 

 tissue PCB concentrations, it was found that ca. 40% of the GC 

 chromatograms (or 67 samples) showed an anomalous peak near the 

 end of the run (Fig. 8) . Contamination of the samples has been 



