PARENT SESSION
MA8 - Ecological Aspects of PCB Interactions
Chair: Gendusa, Tony1, 1 20200 Wambli Lane, Huson, Huson, USA
Co-chair: Zwiernik, Mathew2, 2 224 National Food Safety and Toxicology Blvd., East Lansing, MI, USA
8:00 AM to 12:00 PM - Monday, 18 November 2002
Room Ballroom C
(231) Temporal PCB concentration trends with time varying decay rates in carp and smallmouth bass fillets from the Kalamazoo River and Portage Creek, Mi.
Kern, John*,1, King, Todd2, Ronald, French2, Von Gunten, Brian3, 1 Kern Statistical Services, Inc., Pullman, WA, USA2 CDM, Detroit, MI, USA3 CDM, St. Louis, MO
ABSTRACT- Fish fillet samples were collected from the Kalamazoo River and Portage Creek since the early 1980s to evaluate the human health risks associated with consumption. The National Research Council has proposed temporal reduction of fish tissue concentrations as an indicator of the effectiveness of natural attenuation, because fish consumption is often the pathway of human exposure to PCB. The NRC also pointed out that declining concentration may be due to a combination of source controls as well as natural attenuation. Natural attenuation rates are often expressed as a first order decay rate. However, as immediate effects of source controls are realized and natural attenuation becomes the dominant factor affecting reductions, it would be expected that decay rates would slow with time, invalidating the first order decay assumption. The mixed order model (dC/dt = -kCb) has been proposed as a more flexible alternative for evaluating attenuation rates because the first order assumption is relaxed. As a consequence the decay rate may change with time. We used the mixed order model to evaluate current and future PCB concentration in carp and smallmouth bass from the Kalamazoo River and Portage Creek. Temporal trends in wet-weight PCB concentration were found to be inadequate indicators of natural attenuation because PCB concentration co-varied with lipid-fraction and fish length. To correct for this potentially confounding factor, we used analysis of covariance to adjust PCB concentration for lipid and length effects before fitting the MO model. After adjusting for lipid and length covariation, PCB trends were found to vary temporally and among species and sampling stations within the superfund site. At some sites, decay rates were found to be slower than would be expected under the first order decay assumption.
Key words: polychlorinated biphenyls, natural attenuation, time trend, kalamazoo river