Runoff and persistence of selected herbicides were studied on four small Piedmont watersheds in Georgia during four growing seasons. Seasonal runoff losses were determined relative to watershed management, herbicide type and persistence, mode of application, and time of runoff in relation to application timing. Seasonal losses were usually less than 2% of the application, unless large runoff volumes were generated shortly after application, unless large runoff volumes were generated shortly after application. Average storm herbicide concentrations in runoff were correlated with herbicide concentrations at the 0- to 1-cm depth increment of the watershed soils at the time of runoff. Equations describing soil-based herbicide transfer to runoff were power functions with exponents near unity. Simple relationships such as those developed in this study along with hydrology and erosion/sediment models may be useful in predicting pesticide runoff potential when assessing relative impacts of management decisions.; Runoff and persistence of selected herbicides were studied on four small Piedmont watersheds in Georgia during four growing seasons. This is part of a study designed to provide data for developing and testing mathematical models for agricultural chemical transport. Seasonal runoff losses were determined relative to watershed management, herbicide type and persistence, mode of application, and time of runoff in relation to application timing. Seasonal losses were usually <2% of the application, unless large runoff volumes were generated shortly after application. Average storm herbicide concentrations in runoff were correlated with herbicide concentrations at the 0- to 1-cm depth increment of the watershed soils at the time of runoff. Paraquat concentrations in runoff (predominantly sediment associated) were well correlated and positive with the product of soil herbicide concentration and sediment in runoff. Equations describing soil based herbicide transfer to runoff were power functions with exponents near unity with the form: Y= ax(b). Simple relationships such as those developed in this study along with hydrology and erosion/sediment models may be useful in predicting pesticide runoff potential when assessing relative impacts of management decisions. The next step would be to describe key management practices in terms of these coefficients and exponents.
