Lethality of PAH-Contaminated Sediments to Hyalella azteca is Better Predicted from Freely-Dissolved Concentrations Measured in Interstitial Water Than by Concentrations in Sediment

摘要

Background. Several recent studies have supported an increasing consensus that freelydissolvedconcentrations of HOCs in interstitial water better reflect their bioavailabilitythan do concentrations in sediment. Sediment concentrations coupled with partitioningmodels have traditionally been used to estimate freely-dissolved concentrations, but suchmodels typically overestimate dissolved concentrations by up to two- or three-orders ofmagnitude when applied to historically-contaminated field samples [1-2]. Application ofa “two-carbon” partitioning model (which incorporates natural organic carbon and anthropogenic“black carbon” or “BC”) appears to improve the ability to predict toxicity insome site-specific analyses, but has not been as successful when applied to aggregate datafrom multiple sites [1-2].The alternative to sediment-based partitioning calculations is direct measurement offreely-dissolved PAH concentrations in interstitial water, which eliminates the uncertaintysurrounding partitioning phases and coefficients. Using the theory underlying EPA’ssediment benchmark for PAH mixtures [3], freely-dissolved concentrations of 18 parentand 16 groups of alkyl PAHs (PAH-34) measured in interstitial water can be used to predictthe potency of PAH exposures to benthic organisms like the amphipod, Hyalellaazteca. A standardized method developed by NOAA for measuring PAH-34 in sedimentshas been routinely available for more than a decade, and a method for measuring theirfreely-dissolved concentrations has recently been accepted by the ASTM as methodD7363. Since developing the method for freely-dissolved PAH-34, we have conducted athorough evaluation of both direct measurement of freely-dissolved PAH-34 in interstitialwater and estimation of interstitial water concentrations from sediment concentrationsusing default Koc values [4]. The data include ca. 264 sediment samples from 12 differentsites (primarily pyrogenic or background sources of PAHs) with measurements ofsediment PAH-34, freely-dissolved PAH-34, and H. azteca 28-day survival, as well asinput carbon parameters (TOC and anthropogenic BC) used in popular models to predictfreely-dissolved concentrations from sediment concentrations. Regardless of whichsediment/water partitioning models were used, sediment PAH-34 data were much poorerat predicting mortality than the directly measured freely-dissolved concentrations inwater and the water/lipid partitioning coefficients from EPA [4]. Based only on themeasured freely-dissolved PAH-34 concentrations in interstitial water (and without