and small fish, testing of the model was performed by comparing model results to observed PCB homologue concentrations in white perch. All parameters for this evaluation were previously specified except for the gill transfer efficiency coefficient (p). which was adjusted to 0.25 for simulation results presented below.

A good comparison of model results to observations was obtained for di- through hexa-CB concentrations in white perch at Km 239 (RM 148.5) (see Fig. 25.6 for di-, tri- and penta-CB comparisons) and Km 191 (RM 118.5) (not shown). Di-CB accumulations in perch are quite low (approx. 5 |g g-1 (lipid)) and appear to rapidly adjust to variations in PCB exposure concentration in this portion of the river (for example, see dissolved PCB concentrations in Fig. 25.5). In contrast, accumulations of higher chlorinated homologues in perch are greater (ranging from 10 to 60 |g g-1 (lipid)). This is largely due to increased hydropho-bicity (as represented by the increased Kqw value) of the higher chlorinated homologues which favor their accumulation in the lipid of fish. Accumulation of the more-chlorinated homologues by perch show a clear increase in the early 1990s (corresponding to increased PCB loads from the Upper Hudson). Higher frequency variations that are apparent for dissolved PCB concentrations (see Fig. 25.5) and for di-CB in perch (Fig. 25.6), however, are largely attenuated. This is due to the relatively

0 0

Post a comment