A probabilistic model for predicting distributions of PAH ratios between oysters and marine sediments

Modeling the uptake of organic chemicals by aquatic organisms is an importa nt step in assessing human exposure to harmful environmental contaminants v ia seafood consumption. In many cases, investigators have applied the Equil ibrium Partitioning (EqP) Model to determine the bioavailability of organic contaminants. This model assumes that a thermodynamic equilibrium exists b etween local environmental phases. It relies on point-estimated biota-to-se diment ratios (BSRs) to predict the concentration of organic residuals in t he lipid tissues of organisms from the measured amounts of contaminants in the organic carbon component of the sediment. The paper presents a modeling analysis of polycyclic aromatic hydrocarbon (PAH) oyster and sediment cont amination data collected at Murrells Inlet, SC, a high salinity estuary loc ated in a heavily urbanized area just south of Myrtle Beach, SC. Three repr esentative PAH analytes are considered: phenanthrene (PHE), pyrene (PYR) an d chrysene (CRY). Statistical tests and scatter plots of lipid and organic carbon-normalized data clearly indicate that point-estimated BSRs of the Eq P model are not appropriate for this system. As an alternative, the paper p roposes a probabilistic model that describes the distributions of BSR value s for each analyte. BSR upper limit confidence intervals given by these mod els are equal in magnitude to those derived from equilibrium partitioning. However, contrary to the predictions of the EqP model, BSR values appear to decrease with an increase in the molecular weight of the analyte. This sug gests that heavier PAHs are 'falling out' of the water column into the sedi ment and become less available for uptake by oysters. (C) 2000 Elsevier Sci ence B.V. All rights reserved.