The correlations between the concentrations of polynuclear aromatic hydrocarbons (PAHs) and both 

 amphipod survival and microbial bioluminescence were very strong and highly significant (Table 20). 

 The PAHs showed consistently negative correlations with these end-points in sharp contrast with the 

 correlations with metals and chlorinated compounds. The correlations with the sum of the low molecu- 

 lar weight (2- and 3-ring) PAHs were particularly significant. Also, the concentrations of petroleum- 

 related compounds and microbial bioluminescence were highly correlated. However, the correlations 

 between the bivalve test results and the PAHs were very weak and frequently not significant. Bivalve 

 survival was correlated with the concentrations of only five low molecular weight compounds. Ac- 

 counting for the number of variables (32), only those correlations shown with "**" or "***" would 

 remain significant. 



National sediment quality criteria (SQC) have been proposed for three aromatic hydrocarbons (U.S. 

 EPA, 1994): fluoranthene, acenaphthene, and phenanthrene expressed in units of organic carbon. The 

 correlations between these three compounds normalized to TOC content and both amphipod survival 

 and microbial bioluminescence were significant (Table 20). The correlation between amphipod sur- 

 vival and acenaphthene improved when the chemical concentrations were normalized to the TOC con- 

 tent (Rho = -0.595** vs. Rho =-0.641***). Otherwise, normalization to TOC content tended to dimin- 

 ish the correlative strength between the concentrations of these three compounds and amphipod sur- 

 vival and microbial bioluminescence. 



Table 20. Spearman-rank correlations (Rho, corrected for ties) between four toxicity end-points 

 (as percent of controls) and the concentrations of PAHs in Hudson-Raritan Estuary sediments 

 (n=38). 



82 



