Herbicide transport in a managed riparian forest buffer system

The effect of a riparian forest buffer system (RFBS) on transport of two herbicides, atrazine and alachlor, was studied during 1992-1994. Herbicides were applied to an upland corn crop in March of each year. The buffer system was managed based on USDA recommendations and averaged 50 m in width. The system included a grass buffer strip immediately adjacent to the field (Zone 3); a managed pine forest downslope from the grass buffer (Zone 2); and a narrow hardwood forest containing the stream channel system (Zone 1). After the first year of the study, the managed forest was clear-cut in 1/3 and thinned in 1/3 of Zone 2. The other 1/3 of Zone 2 was left as mature forest. Most of the herbicide transport in surface runoff occurred before 30 June with about 25 cm of cumulative rainfall after herbicide application. During this period of higher herbicide transport, atrazine and concentrations averaging 34.1 micrograms L-1 and 9.1 micrograms L-1 at the field edge, respectively, were reduced to 1 microgram L-1 or less as runoff neared the stream. There were generally no differences among the mature forest and the two treatment areas (clear-cut and thinned) for either concentration or load in surface runoff. Using precipitation data collected on site, the effects of dilution versus other concentration reduction factor (infiltration, adsorption) was estimated for surface runoff. Concentration reduction was greatest per meter of flow length in the grass buffer adjacent to the field. There was only minor transport of herbicides through the buffer system in shallow groundwater and little difference between the Zone 2 treatment areas. In 1992 and 1993, herbicide concentrations in shallow groundwater in the RFBS and at the edge-of-field were generally at or below detection limits. In 1994, well concentrations of both herbicides increased, probably in response to infiltration of surface runoff containing high herbicide concentrations. Average herbicide concentrations were at or below detection limits in groundwater near the stream for most of 1994.