T10 PM Advances in Bioaccumulation Assessment
Tuesday, 15 November 2005: 1:50 PM - 5:30 PM in 341-342

(ROS-1117-725988) Reducing uncertainties in biotransfer modeling in meat and milk.

Rosenbaum, R¹, McKone, T², Jolliet, O¹, ¹ Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland² University of California, Berkeley, California, United States of America

ABSTRACT- Improving indirect (human) exposure assessment with respect to biotransfer into biota is the main objective of this study. Much of the overall uncertainty in exposure is attributable to the estimation of biotransfers. Biotransfer factor (BTF) and carry over rate (CR) are measures of the fraction of ingested contaminant transferred to animal tissue. The most commonly used BTF model dates back seventeen years and in spite of its widespread use in multimedia exposure models few attempts have been made to advance or improve the highly uncertain Kow regressions. Furthermore, in the range of Kow where meat and milk become the dominant human exposure pathways these models provide clearly unrealistic CRs. Quantitative structure-property relationships, such as the molecular connectivity index (MCI), derived from the structural formula of a chemical and then correlated to experimentally determined BTFs/CRs, are explored for possible combination with an improved Kow regression further decreasing indirect exposure uncertainty. Literature suggests that MCI-methods significantly increase the accuracy of BTF/BCF estimations, reducing at the same time uncertainties linked to unreliable Kow measurements. A dynamic six compartment cow model has been developed to track the transport of a chemical into meat fat and milk within the cow. First results demonstrate that steady state is almost never reached in meat which in some cases also holds true for milk. Degradation within animal tissue is a currently neglected but important fate process to be included. Another problem of current biotransfer models was identified as they typically adopt single measured data points which often do not represent steady state concentrations, as the feeding experiment duration was not long enough. The dynamic cow model can assist in reinterpreting experimental data, including predicting the extension of experimental time series and the according (non-steady state) concentration reached during the cow lifetime. This leads to more consistent, less varying data points to derive a less uncertain CR-prediction regression model, combining Kow and MCI.

Key words: biotransfer, cow, milk, meat