Doctor of Philosophy (Ph.D.)
Virginia Institute of Marine Science
Surrogate contaminant variables (heavy metals, organics, physical oceanograpic data) can be used to predict the incidence of positive cultures of Acanthamoeba sp. in oceanic sediments. Amoebae data are drawn from five years of study involving stations in Narragansett Bay, Rhode Island, the New York Bight, and the Philadelphia-Camden dumpsite, and associated pollution parameters are drawn from literature sources, computerized marine pollution data bases, and other archives. The Statistics Analysis System (SAS) MAXR('2) improvement technique (stepwise regression) and general linear model procedures are used to generate correlations for surrogate variables and produce final predictive models and tables. Model procedures for the three study areas are most valid for Narragansett Bay and the New York Bight but less valid for the Philadelphia-Camden dumpsite due to the small quantity of relevant data. The Durbin-Watson statistic is used to test for autocorrelation of model residuals and, using this test, the Philadelphia-Camden model is again found to be the least valid, although applicable within limits. The division of contaminant variables into "tactical" (short-term, simple analysis) and "strategic" (long-term, more complex analysis) categories enhances the predictive effort through the introduction of a cost-effective procedure evaluation. Generally, the simple variables predict the incidence of positive Acanthamoeba cultures as well as the more complex data sets. There are sufficient data and applicable computer programs to produce useful results for an investigation involving potentially pathogenic protozoans and public health management decisions may be made using the tables and formulae generated using these procedures.
© The Author
Berman, Carl Robert Jr, "A statistical model to predict the incidence of pathogenic protozoa (amoebida: acanthamoebidae) in oceanic sediments using surrogate variables" (1983). Dissertations, Theses, and Masters Projects. William & Mary. Paper 1539616564.