Testing the Virtual Beach Escherichia coli Prediction Model on Sugar Island, MI

Testing the Virtual Beach Escherichia coli Prediction Model on Sugar Island, MI

For over a third of the last three summers, beach water along the North Shore of Sugar Island, MI has been under swimming advisories due to fecal contamination. Currently, the State of Michigan mandates an 18 hour culture test be used to measure the E. coli bacteria levels that can often change by the hour. One of the ways that beach managers can avoid this time delay is to use models that can predict bacteria levels using regression equations, fitting lines to the mechanisms that cause these beach closures. The models include environmental variables such as precipitation, wind, and temperature. The created Virtual Beach models for Sugar Island were measured by two criteria. First, how the models fit linearly to environmental variables and second, how often did the models correctly predict dangerous levels of E. coli and how many times did the model incorrectly predict E. coli levels greater than 300 CFU/100 ml during times of safe bacteria amounts. Both models that were created for the North Shore failed to predict most of the dangerous E. coli levels in the beach water. The model with the most data had more favorable correlations with environmental data. More variables should be added to these models to best predict the harmful bacteria levels as the mechanisms that cause beach closures on Sugar Island are not yet understood.

Albrecht, Shane

Biology

Lake Superior State University

2011

Copyright Shane Albrecht: All rights reserved. LSSU use only.

application/pdf

English

text.monograph

S20230905009