Recommendations for the derivation and use of biota–sediment bioaccumulation models for carcinogenic polycyclic aromatic hydrocarbons

Abstract

Abstract Carcinogenic polycyclic aromatic hydrocarbons (cPAHs) are important sediment contaminants that can pose health risks to people who eat shellfish from contaminated sites. Biota–sediment accumulation factors (BSAFs) are quotients of colocated lipid-normalized tissue concentrations and organic carbon (OC)–normalized sediment concentrations, whereas biota–sediment accumulation regressions (BSARs) are models describing the relationships between these tissue and sediment concentrations. BSAR/Fs (BSARs and/or BSAFs) are commonly used to back-calculate sediment preliminary remediation goals (PRGs) from target tissue concentrations; the PRGs are then used to set target action levels (i.e., sediment concentrations above which remedial actions will be prescribed). The cPAH BSAR/Fs reported across sites and species are highly variable due to both site- and species-specific differences and inconsistent BSAR/F calculation methods and assumptions. We reviewed past studies, identified best practices for developing BSAR/Fs, and compiled publicly available colocated tissue and sediment data for 7 cPAHs from 13 sites across the United States. Of the 249 unique cPAH data sets compiled for various species, only 17 yielded acceptable BSAR/Fs, 16 of which were for clams. The influence of BSAR/Fs on sediment remedial action decisions and costs can be disproportionate to the quality of the statistical models from which they are derived. Therefore, it is important to establish and follow best practices for deriving BSAR/Fs and for deciding whether and how BSAR/Fs should be used. Based on our review and analysis, we highlight the advantages of relying on BSARs and propose a consistent method for deriving and judging the reliability of these relationships. We also offer guidance for evaluating the ramifications of BSAR uncertainty on remedial decision making at contaminated sediment sites, and we discuss alternative ways to make risk management decisions in the absence of a reliable site-specific BSAR. Integr Environ Assess Manag 2017;13:1060–1071. © 2017 SETAC Key Points As compared with biota-sediment accumulation factors (BSAFs), biota-sediment regression models (BSARs) are advantageous because they allow for a better evaluation of relationship uncertainty and because they can account for the impact of non-sediment factors (e.g., water) on tissue concentrations. When developing a BSAR, the data should be obtained from well-designed sampling programs and evaluated using a consistent, defensible methodology that includes an assessment of the reliability and uncertainty in the observed tissue–sediment relationship. If a suitable relationship cannot be developed using site-specific data, or if the level of uncertainty associated with BSAR/Fs is unacceptably high to inform remedial decision making, costly remedial decisions should not continue to be based on invalid BSARs. When an acceptable BSAR cannot be developed, other approaches include basing decisions on direct contact scenarios, using adaptive management techniques, collecting more data to improve the BSAR, or using a mechanistic bioaccumulation model.