SPATIAL AND TEMPORAL DYNAMICS OF VIBRIO PARAHAEMOLYTICUS IN THE CHESAPEAKE BAY
Embargo until
Date
2017-08-21
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Johns Hopkins University
Abstract
Vibrio parahaemolyticus is one of the leading causes of seafood-associated illnesses in the United States. Raw shellfish are the primary exposure route, as the bacterium is naturally prevalent in estuarine environments where shellfish are commonly harvested. Climatic changes such as ocean warming are predicted to increase the prevalence of Vibrio spp. globally. There is therefore a need to better predict and quantify the health risks of V. parahaemolyticus in order to mitigate the public health burden of this bacterium.
This dissertation set out to better characterize the spatial and temporal dynamics of V. parahaemolyticus in an estuarine environment, specifically the Chesapeake Bay, through three aims. The first aim was to develop and evaluate new geostatistical prediction methods that ensured mathematical validity when using non-Euclidean distances. New and existing methods were compared using cross-validation techniques across multiple scenarios. The second aim was to identify and quantify the environmental determinants of V. parahaemolyticus in the Chesapeake Bay using interval-censored regressions and accounting for non-linear associations. The third aim was to develop and evaluate prediction and forecast models for the bacterium using time-indexed and –lagged water quality measures using Tobit regressions and cross-validation metrics.
Results for Aim 1 identified a balanced tradeoff of prediction accuracy and prediction variance for one of the new methods, which overall outperformed existing methods. Aim 2 confirmed previously reported environmental associations with V. parahaemolyticus and also identified new and more nuanced associations concerning water temperature, salinity, and turbidity. Results for Aim 3 identified multiple prediction/forecast models that performed adequately, and found that lagged environmental measures contained useful predictive power for these models.
Overall, this dissertation imparts new methods for spatial predictions, improves the understanding of the V. parahaemolyticus ecological paradigm, and indicates that useful predictions for this bacterium can be developed using only in situ water quality measures. This work has contributed significantly to the fields of spatial statistics, microbial ecology, and food safety forecasting. The culmination of this dissertation also lays the groundwork for future research efforts involving space-time predictions of V. parahaemolyticus as well as the quantification of health risks to shellfish consumers.
Description
Keywords
Vibrio parahaemolyticus, Chesapeake Bay, geostatistics