Theses and Dissertations

Issuing Body

Mississippi State University


Shaw, David R.

Committee Member

Massey, Joseph H.

Committee Member

O’Hara, Charles G.

Committee Member

Bingner, Ronald L.

Committee Member

King, Roger L.

Date of Degree


Document Type

Dissertation - Open Access


Weed Science

Degree Name

Doctor of Philosophy


College of Agriculture and Life Sciences


Department of Plant and Soil Sciences


Little historical water quality data is available for the Upper Pearl River Basin (UPRB), yet there are UPRB waters listed as impaired. Objectives of this research were to measure pesticide and sediment concentrations in UPRB surface waters and validate the Annualized Agricultural Nonpoint-Source (AnnAGNPS) runoff model with the measured data for a portion of the UPRB. An additional objective was to quantify effects of land use changes on UPRB surface waters from 1987 to 2002 using AnnAGNPS. Of the fifteen compounds analyzed, hexazinone was most frequently detected, in 94% of samples, followed by metolachlor, tebuthiuron, and atrazine. Metribuzin was detected in only 6% of samples. Total dissolved solids (TDS) concentrations were highest at Carthage, which drains the largest area of three sites sampled for TDS. Most samples measured below Environmental Protection Agency (EPA) standards for pesticides and TDS in drinking water and also below levels toxic to aquatic organisms. For eight of twelve months analyzed between October 2001 and January 2003, average monthly sediment loadings for measured and AnnAGNPS-simulated data differed no more than 109%, resulting in an R&178; value of 0.328. A comparison of measured and simulated atrazine and metolachlor loadings by event resulted in R&178; values of 0.095 and 0.062, respectively. Most daily atrazine and metolachlor loadings for measured and predicted data were very low. On May 18, 2003, AnnAGNPS predicted a metolachlor loading of 80 mg, while measured data showed a loading of 5.6 mg. Measured data showed an earlier spike on January 20, 2003 that was not mirrored by the model. Atrazine comparisons followed the same trend, except measured loadings did not spike until February 22, 2003. The 2002 AnnAGNPS simulation resulted in 15% more average annual runoff than the 1987 simulation, although both simulations had the same precipitation. The 2002 simulation also had higher values for sediment and organic carbon loading. Nitrogen loading was the only runoff or pollutant loading category that was less for 2002 than for 1987. Urban land cover contributed more runoff and pollutant loadings from 1987 to 2002, while traditional row crop agriculture had less of an impact on pollutant loadings.