Derivation of shellfish harvest reopening criteria following the New Carissa oil spill in Coos Bay, Oregon

Oil spills in Alaska, California, Maine, and other states have raised conce rns regarding potential contamination of fish and shellfish, and have led t o temporary closures of seafood harvests while health risks are assessed. L acking standardized protocols, these assessments are generally ad hoc, site -specific efforts, with significant differences in risk evaluation criteria . This article describes the response of a !:tate health agency to shellfis h contamination following an oil spill on the Oregon coast, and discusses s ome of the factors that can complicate the evaluation of potential health r isks from consumption of oil-contaminated shellfish. On 4 February 1999, th e Japanese-owned cargo ship M/V New Carissa, carrying an estimated 400,000 gallons of light diesel and heavy fuel oil, ran aground 2 miles north of Co os Bay, Oregon. Damage to the ship's hull from the grounding and pounding s urf caused the release of an estimated 25,000 to 70,000 gallons of oil. Con cern for potential contamination of local recreational shellfish and commer cial oyster beds prompted the Oregon Department of Agriculture (ODA) to clo se shellfish harvesting in Coos and Douglas counties. ODA requested assista nce from the Oregon Health Division in the derivation of risk-based criteri a for reopening the shellfish harvest. Criteria were developed for the prim ary contaminants of concern, polycyclic aromatic hydrocarbons (PAHs) expres sed as total benzo[a]pyrene (BaP) equivalents. "Safe" (<10 mu g/kg) and "un safe" (>4.5 mu g/kg) BaP equivalent levels were derived based on upper end (32.5 g/d) and average (7.5 g/d) estimates of shellfish consumption, respec tively. Composite samples of oysters, clams, and mussels (15-20 per composi te) were collected from target ai eas and analyzed for PAHs by gas chromato graphy/mass spectroscopy (GC/MS). Carcinogenic PAHs were converted to total BaP equivalents (wet weight) and compared with criteria. Two oyster sample s, collected from a slough off of Coos Bay, contained 33.9 and 34.5 mu g/kg BaP equivalents; ail other samples had less than 10 mu g/kg BaP equivalent s. An evaluation of the PAH profiles in the two higher oyster samples indic ated a primary source other than the New Carissa oil. Because shellfish sam ple BaP equivalents attributable to the New Carissa oil spill did not excee d 10 mu g/kg, shellfish harvesting was reopened on 4 March 1999. This study revealed some of the inherent difficulties in attempting to quantify healt h risks from contaminated shellfish following an oil spill and demonstrated the clear need for standardized protocols for responding to such events.