A spatially and temporally variable model of mercury concentrations in aquatic communities with applications to public health protection and water quality assessment
Abstract
Mercury (Hg) is an important environmental contaminant capable of gravely impacting human health through consumption of contaminated fish. In recognition of this fact, many health and environmental agencies spend large amounts of time, effort, and money collecting and chemically analyzing fish tissue samples in order to provide the public with fish consumption recommendations, identify contaminated sites, and assess temporal trends of contaminant concentrations. In this research, a comprehensive observational data analysis method for monitoring Hg concentrations in aquatic communities is developed and demonstrated that accurately performs these applications. The proposed method also minimizes sample replication requirements to reduce monitoring costs. A data set of 915 samples from 268 sites (353 sampling events) in Indiana is used to test this model. The analysis of Indiana's data set estimates a decline in statewide biotic-Hg concentrations of 23% between 1983 and 1991. The majority of this decline is thought to be due to reductions in regional-scale atmospheric Hg deposition rather than changes at the global or local (point source) scale. Additionally, this analysis indicates flood control reservoirs are a significant source of Hg contamination in the waterways that receive the discharge of these reservoirs. Lastly the three applications are demonstrated using this model to predict fish tissue concentrations, measure deviations from background conditions, and perform valid hypothesis testing.
Degree
Ph.D.
Advisors
Spacie, Purdue University.
Subject Area
Biostatistics|Environmental science|Toxicology|Public health|Freshwater ecology
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.