MA2 Exposure to POPs Through Terrestrial Vegetation
255 Portland Ballroom
8:00 AM - 12:00 PM, Monday
() Re-evaluation of Kow-based bioaccumulation models for plants.
Sample, B1, Burris, J2, Loveridge, A1, Tossel, R1, 1 CH2M HILL, Sacramento, CA, USA2 Syracuse Research Corp, Denver, CO, USA
ABSTRACT- Estimation of dietary contaminant exposure through plants for humans and wildlife requires measured or estimated concentrations. When measured data are absent, the Kow-based model for plant foliage developed by Travis and Arms (1988) is frequently used. An extensive review of the model and underlying data, conducted as part of the EPA EcoSSL effort, identified several issues affecting the usability of the model. Because of these issues, a re-evaluation of the Travis and Arms model was undertaken. All literature cited by Travis and Arms was acquired and suitable bioaccumulation data were extracted. Data from more recent studies were also included to increase the representation of some chemical classes (i.e., polycyclic aromatic hydrocarbons [PAHs] and polychlorinated biphenyls [PCBs]) and to expand the overall dataset. A series of regression models relating Kow to BAFs and analyte-specific soil-plant tissue models were developed. These models were compared to the Travis and Arms model, uncertainies and limitations of these new models, and their recommended applications were made. Bioaccumulation data were extracted from 27 studies: 17 previously cited in Travis and Arms plus 10 additional studies. These data represented over 40 chemicals in nine chemical classes (benzophenyl ureas, carbamates, cyclodienes, DDT and metabolites, nitro/chlorinated aromatics, PCBs, PAHs, triazines, and other), four tissue types (i.e., foliage, fruit, roots, and seeds), and 44 plant species. For all tissues except unrinsed fruit, an increase in Kow was associated with a decrease in BAF. All regression models except combined rinsed and unrinsed fruit and only unrinsed fruit had p-values less than 0.05, indicating reliable prediction of BAF with Kow using the new models. Although the Travis and Arms model lies within the 95% confidence intervals of the new model for rinsed foliage, its more negative slope indicates lower predicted BAFs at higher Kows than those based on the new model.
Key words: organic compounds, bioaccumulation, plants, modeling