benthic communities indicated significant differences between all sites. Most of the 

 cytochrome P-450 end-points and two of the micronuclei end-points indicated differences 

 between sites, and, therefore, rated yes's for the "sensitive" criterion. 



Half of the sediment toxicity end-points rated a "yes" for the "correlated with 

 toxicants" criterion, including those of A. abdita survival, D. gyrociliatus reproduction and S. 

 purpuratus echinochrome content and cytogenetic/mitotic abnormalities. None of the fish 

 tests were very highly correlated with the concentrations of chemicals in the fish livers. 

 Since no benthos data are available for the most contaminated site, correlation analyses 

 have not yet been performed. 



Again, half of the sediment toxicity end-points rated a "yes" for not being correlated 

 with nuisance variables (grain size and TOO. The benthos data have not yet been tested 

 for correlations with sedimentological factors. The AHH, micronuclei, steroid hormone, and 

 most of the P-450 end-points were not correlated with length, weight, or organ weight of the 

 fish. However, HSI was inversely correlated with P-450E content and P-450E/total P-450; 

 while estradiol 2-hydroxylase was positively correlated with GSI. 



The two avoidance end-points in the amphipod tests had very high analytical 

 variability and rated a "no" in Table 43 for the "low analytical variability" criterion. The 

 average CVs of two of the S. purpuratus cytogenetic/ mitotic end-points were very high, 

 while those of two others were relatively low. One or more of the AHH, P-450 and hormone 

 end-points had relatively high variability among fish at the sites while the others had 

 low variability. All of the micronuclei end-points had relatively high variability among 

 fish at sampling sites. 



The avoidance end-points in the amphipod tests and some of the S. purpuratus 

 cytogenetic/mitotic end-points had relatively low discriminatory power. R. abronius 

 survival and M. edulis percent normal end-points had the highest values among the 

 sediment toxicity tests. Discriminatory power of the AHH end-points was relatively low. 

 The discriminatory power of most of the P-450 end-points exceeded a value of 2.0. The 

 testosterone concentration end-point had a relatively high discriminatory power, while that 

 of estradiol content did not. The discriminatory power of the detached micronuclei end-point 

 exceeded 2.0, whereas those of total and attached micronuclei did not. 



94 



