Determining Historical Pesticide Deposition on Cape Cod through Sediment Core Analysis:A Validation of GIS as An Exposure Assessment Tool

Geographic Information Systems (GIS) has emerged as a powerful tool to assess current and historical exposure to environmental pollutants. GIS aids in the visualization and understanding of associations between exposure to contaminants and disease. This study is an example of the bridge between environmental science and public health and of how new technology such as GIS can be incorporated into these fields to strengthen both the research and the communication of scientific results. It attempts to validate a GIS-based aerial drift model which predicts the residential exposure to and boundaries of historical organochlorine pesticide (OCP) drift from applications on cranberry bogs, tree pest sprayings and others by analytically quantifying the historical pesticide deposition in a transect of lakes radiating from a distinct spray source. This model was previously used to assess historical residential exposure to OCPs in an environmental epidemiological case-control study of breast cancer incidence on Cape Cod, MA, where the incidence rate of the disease is significantly higher than in the rest of the state. The model's validation in this current study is essential to establishing its predictive ability and thus, its further use. Ground truthing of the model was done through the collection and analysis of sediment cores along a transect of five hydrologically independent kettle ponds radiating from a distinct OCP tree-pest spray area. Measurements of OCP concentrations, total carbon and total organic carbon were determined, and dating of the sediments was completed using 210Pb and verified using 137Cs. Each 50-cm core was sliced into 25 2- cm sections for the analyses, creating a fine-scale depositional history in each pond. Information gathered from each core allows for the determination of the extent and degree of dissipation of individual spray events of a known source area and determine how well the model fits the actual data.