tions of parent 4,4'-DDT in the highly toxic samples exceeded the average concentration in the non- 

 toxic samples by a factor of 20.3 and exceeded the ERM value of Long and Morgan (1990) by a factor 

 of 21.6. The other isomers of DDT were highly elevated in the highly toxic samples and often ex- 

 ceeded respective ERM guideline values. The sums of the quantified PCB congeners were multiplied 

 by 2.0 to estimate the concentration of total PCB (NOAA, 1989). The average concentrations of total 

 PCBs in the highly toxic samples exceeded the average concentration in the nontoxic samples by a 

 factor of 1.9 and exceeded the ERM value of Long et al. (1995) by a factor of 4.1. 



The concentrations of all categories of PAHs were considerably elevated in the samples that were 

 highly toxic to the amphipods relative to the samples that were not toxic (Table 24). The concentra- 

 tions of organic carbon, inorganic carbon, and fine-grained sediment particles were not elevated in the 

 highly toxic samples to the same degree as the PAHs. The average concentrations of total low molecuar 

 weight PAHs in the toxic samples (34,672 ppb) exceeded the average concentrations in the nontoxic 

 samples (922 ppb) by a factor of 37.6 and exceeded the ERM value for LPAH of Long et al. (1995) by 

 a factor of 11.0. Also, both the high molecular weight compounds and total PAHs were elevated in 

 concentration in the toxic samples relative to the nontoxic samples. The concentrations of both 

 fluoranthene and phenanthrene in the highly toxic samples exceeded both the average concentrations 

 in the nontoxic samples and the respective SQC concentrations by a considerable amount. Although 

 the average concentration of acenaphthene in the toxic samples exceeded the nontoxic average by a 

 factor of 58.7, it exceeded the SQC by a factor of only 2.8. 



In tables 25-27 the average concentrations of chemicals in samples that were toxic to microbial bi- 

 oluminescence were compared with those that were not toxic. As observed in the amphipod tests, the 

 average trace metals concentrations were relatively similar in the toxic and nontoxic samples, as indi- 

 cated by ratios between the averages of 1.0 or therabouts (Table 25). Among the metals that were 

 quantified, the concentrations of zinc were most elevated in the highly toxic samples; the average 

 concentration of 442 ppm in the highly toxic samples exceeded the average in the nontoxic samples 

 (240.7 ppm) by a factor of 1 .8. Also, the concentration of lead in the highly toxic samples (average of 

 224 ppm) exceeded the average concentration in the nontoxic samples (132.4 ppm) by a factor of 1.7. 

 The average concentrations of both lead and zinc in the highly toxic samples were very similar to the 

 ERM values (218 and 410 ppm, respectively). The average concentrations of mercury in all three 

 categories were very similar (2.0-2.4 ppm) and exceeded the ERM value of 0.71 ppm. The concentra- 

 tions of trace elements simultaneously extracted with the acid-volatile sulfides were very similar among 

 the three toxicity categories. The SEM/AVS ratios averaged 1 .0 in the nontoxic samples and 0. 1 in the 

 significantly toxic and highly toxic samples. 



Again, as observed in the amphipod tests, most of the pesticides and other chlorinated organics oc- 

 curred in similar concentrations in both the toxic and nontoxic samples (Table 26). The concentrations 

 of some compounds, such as heptachlor were below the detection limits in all samples, as indicated by 

 standard deviations of 0.0 in all categories. The average concentrations of many compounds (e.g., 4,4'- 

 DDE, 4,4'-DDT) actually were considerably lower in the highly toxic samples than in the nontoxic 

 samples, despite the significant Spearman-rank correlations observed with these data. However, the 

 concentrations in the highly toxic samples often exceeded the respective ERM values. The average 

 concentrations of total PCBs were relatively high, exceeding the ERM value of 180 ppb, in all catego- 

 ries. One highly toxic sample from site 12 in the East River had a detectable concentration of hep- 

 tachlor epoxide, thus driving up the average for that compound. 



91 



