Toxic Elements in Aquatic Sediments: Distinguishing Natural Variability from Anthropogenic Effects
2009
Hou, Aixin | DeLaune, R. D. | Tan, MeiHuey | Reams, Margaret | Laws, Edward
Regressions of aluminum against potentially toxic elements in the sediments of freshwater aquatic systems in Louisiana were used to distinguish natural variability from anthropogenic pollution when elemental concentrations exceeded screening effects levels. The data were analyzed using geometric mean model II regression methods to minimize, insofar as possible, bias that would have resulted from the use of model I regression. Most cadmium concentrations exceeded the threshold effects level, but there was no evidence of an anthropogenic impact. In Bayou Trepagnier, high concentrations of Cr, Cu, Pb, Ni, and Zn appeared to reflect anthropogenic pollution from a petrochemical facility. In Capitol Lake, high Pb concentrations were clearly associated with anthropogenic impacts, presumably from street runoff. Concentrations of potentially toxic elements varied naturally by as much as two orders of magnitude; hence it was important to filter out natural variability in order to identify anthropogenic effects. The aluminum content of the sediment accounted for more than 50% of natural variability in most cases. Because model I regression systematically underestimates the magnitude of the slope of the regression line when the independent variable is not under the control of the investigator, use of model II regression methods in this application is necessary to facilitate hypothesis testing and to avoid incorrectly associating naturally high elemental concentrations with human impacts.
Show more [+] Less [-]AGROVOC Keywords
Bibliographic information
This bibliographic record has been provided by National Agricultural Library
