ABSTRACT
In many areas of intensive corn production, atrazine and metolachlor are among the most commonly found herbicides in surface and ground water. This 2-yr study compared the impact of grass and grass+tree buffer strips on the exported masses of atrazine, metolachlor, and a degradation product of atrazine, desethylatrazine (DEA). The experimental system consisted of four replicate plots in a three-way completely randomized design (no buffer zone, grass buffer zone, and grass+tree buffer strips). The field plots were 5 m wide and 30 m long and grown in corn. The grass and grass+tree buffer strips were 5 m and had the same grass vegetation except for eight young hybrid poplars. Over the 2-yr study, surface runoff and subsurface infiltration water (under the buffer strip) were collected after the initial three rainfall events after herbicide application. Dissolved atrazine, metolachlor, and DEA were analyzed by gas chromatography/mass spectrometry. The presence of buffer strips decreased the exported masses of atrazine and metolachlor in surface runoff. A three-way ANOVA with treatment (type of buffer strip), water (surface runoff or subsurface infiltration), and time between herbicide application and rainfall event as factors showed a significant reduction (40-60% in 2004 and 75-95% in 2005) in the total (surface runoff+infiltrated water) exported masses of atrazine and metolachlor in the presence of buffer strips. Rainfall events after herbicide application were different between the 2 yr and greatly affected the flow distribution (e.g., subsurface infiltration) and the leached herbicide concentrations. No significant difference in the capacity to reduce herbicide exports was observed between grass and grass+tree buffer strip treatments; the poorly developed young poplar biomass at the time of the study may partly explain this observation.
