Toxicity to bivalve development was not nearly as high in lower Raritan River sediments (site 20) as in 

 the amphipod test. All three of the stations at site 39 were significantly toxic to amphipods and two of 

 the three were toxic to both end-points of the bivalve larvae test, but the site mean was not significantly 

 different from controls in any of the three test end-points. Two of the three samples from site 30 in 

 Sandy Hook Bay were toxic to all three test end-points and toxicity diminished northeastward out into 

 lower New York Harbor. 



Microbial Bioluminescence Tests of Organic-Extracts. An initial range-finding experiment was 

 conducted with sediments previously tested with the amphipods. Each of the sediments that indicated 

 high, intermediate, and low toxicity to the amphipods were tested with three sediment concentrations 

 (3, 10, and 15 g wet weight of sediment) to provide a dilution series. This experiment showed that 

 extracts from 3 g of sediments were sufficient to cause a 50% reduction in light output by the Microtox tm 

 bacteria. In addition, separation and precipitation of extract phases occurred in the vials during the 

 extraction procedures with the 10 and 15 g extracts from both the intermediate and high toxicity samples. 

 These results indicated that the 10 and 15 g concentrations were too high and would lead to spurious 

 light attenuation. A total of 1 16 of the 117 sediment samples was tested with Microtox tm . The sedi- 

 ment concentrations (EC50s) that caused 50% light inhibition were determined, along with the 95% 

 confidence limits, for the Central Long Island Control and each test sediment. Duplicate tests were run 

 for each sample. Mean EC50 values (mg sediment/mL) and the 95% confidence intervals are listed in 

 Table 9. Samples that were significantly different from controls are listed with one asterisk. Those 

 samples in which the mean EC50 was 80% or less of the control are listed with two asterisks. 



The mean EC50s of the two tests of the controls were 2.02 and 2.1 mg/mL (Table 9). Of the 116 

 samples that were tested, 47 (41%) were significantly toxic (i.e., different from controls) in this test. 

 Many of the samples (32) caused EC50 values of 1.6 mg/mL or less (80% of controls). However, in 

 some cases the test samples were less toxic than the CLIS controls (as indicated by EC50 values greater 

 than 2.1). This test indicated that 19 of the 39 sites (49%) were significantly different from controls. 

 The mean EC50s for 14 sites were significantly different from controls and 80% of the control re- 

 sponse or less. 



All three of the sites in zones C and D were significantly different from controls, whereas none of the 

 sites in zones A, G, and M were toxic in this test. Stations 6-C, 9-B, 28-A, and 36-B were the most 

 toxic, as indicated by the lowest mean EC50s. All but one of the nine stations in zone B (western Long 

 Island Sound), and zone D (lower East River) were different from controls in this test. None of the 

 stations in zone A (lower Hudson River) and zone M (New York Bight) were toxic, and only one each 

 in zone E (Upper New York Harbor) and zone L (Lower New York Harbor) was different from con- 

 trols. 



Table 9. Results of Microtox 1 " 1 tests of microbial bioluminescence in organic extracts of sedi- 

 ments; mean EC50's (n=2) and 95% confidence intervals for stations, and mean EC50's (n=3) for 

 sites. 



Regional Sampling Mean EC50 95% Confidence 



Zone Site/station (m g/mL) Interval 



CLIS Control 1 2.02 2.00-2.03 



2 2.11 2.01-2.16 



Mean 2.06 n/a 



50 



