Can Land-use Patterns Serve as a Predictor of Pesticide Occurrence Within an Urban Landscape?

Abstract

The National Water-Quality Assessment (NAWQA) Program of the U.S. Geological Survey (USGS) has focused on relations between land use and water quality in the Nation's streams. The NAWQA design to assess water-quality conditions is based on monitoring streams located in relatively small watersheds (60-150 square miles) that contain a predominance of a single targeted land use. In some NAWQA study areas, such as the Apalachicola-Chattahoochee- Flint River basin, additional spatial surveys were conducted to evaluate the variability of water-quality conditions within and among watersheds representing each targeted land use. Recently (1996-99), the USGS created a digital land-use and land-cover database for most of the upper Chattahoochee River basin and Metropolitan Atlanta, Georgia. The new land-use data are more detailed and cover a larger area of Metropolitan Atlanta than previously available data. This paper addresses whether land-use patterns obtained from this new digital database may be used to predict pesticide concentrations along a gradient of urban land use. Preliminary analyses indicate that pesticide concentrations in streams increase as the percentage of the associated watersheds that may be treated with pesticides increases. Three classes of pesticides were investigated: selective pre-emergent herbicides, insecticides, and nonselective herbicides. The relation between land use and pesticide concentrations is substantially better for selective pre-emergent herbicides, the most widely used class of pesticides, than for the other classes. Additional explanatory information is needed to improve these relations.