Long-term CREAMS and GLEAMS model simulations were used to assess significant differences in potential pesticide and nitrate movement to groundwater as affected by several combinations of tillage, cropping, and nutrient and pesticide management practices. The study area is in Richmond County in the Coastal Plain region of Virginia. Alternative management scenarios were based on crop management systems common to the area. Average annual loads from 35-year simulations were used in the analysis. Friedman's distribution-free analysis of variance and sign tests were found to be appropriate statistical procedures to assess significant differences between effects of management practices.

Commercial nitrogen split-applied at recommended rates leached less than equivalently applied N from a single pre-plant poultry litter application. Leaching increased, but differences were less, when split-applied commercial N and pre-plant poultry litter applications were increased, with both sources apparently supplying N in excess of crop uptake potential. A 3-way instead of a 2-way split commercial N application reduced nitrate leaching minimally. Though tillage-cropping practices did affect percolation volumes, this did not result in high variability in leaching between practices. Leaching was most sensitive to crop-available N inputs regardless of tillage-cropping practices or the methods of N application.

Pesticides with lower soil-adsorptivity, such as atrazine, had greater leaching losses especially on rotations with higher infiltration. However, a relatively high surface- application rate of atrazine for no-till corn did not produce higher leaching losses because additional losses were mostly by volatilization. Surface losses of highly adsorbed chemicals, such as gramoxone, were substantially reduced from no-till fields where runoff and erosion were reduced. Percent reductions in pesticide application rates resulted in equal or greater percent reductions in loading in surface runoff and percolation.