Spatial Processes and Ecotoxicological Risk Assessment Modeling

This dissertation considers the importance of spatial processes in the application of models for environmental risk assessment. The appropriate incorporation of space as an ecological variable to construct spatially delineated risk estimates can improve estimates of the risk outcomes and improve resulting risk management decisions. Three environmental problems are considered, corresponding to three different case studies measuring an ecotoxicological effect at progressively broader spatial scales. The first application addresses how the presence of spatial auto­correlation can affect the exposure of ecological receptors in the environment. The presence of significant spatial correlation can impact methods for adequately determining summary concentrations, hypothesis testing, and for identifying zones in need of remediation at contaminated sites. The second case study addresses how treating habitat as a spatial variable can impact estimates of population doses. Calculated dose estimates are dependent on the distributions of spatial habitat and the toxin. Risk estimates can differ considerably when compared to the more common use of summary statistics to represent an exposure concentration for risk purposes. Measuring the spatial correlation of habitat quality and contamination distribution is recommended as a means for evaluating whether incorporating habitat and spatial movement models are useful for an environmental assessment. Finally, the risk of insects developing resistance to genetically modified crops is evaluated within a spatial milieu of varying bat predation rates and changing genetically modified crop types and percentages. This model is developed at a broad spatial scale and considers large­scale insect migrations and varying spatial pressure from foraging bats. Incorporation of these larger scale processes significantly impacts temporal estimates of the evolution of insect resistance to toxins in the environment.