Submitted to: Journal of Environmental Quality
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/21/2003
Publication Date: 6/20/2003
Citation: Potter,T.L., Truman,C.C., Bosch,D.D., Bednarz,C. Cotton defoliant losses in surface runoff as a function of active ingredient and tillage practice. Journal of Environmental Quality. 2003. v.32:2180-2188. Interpretive Summary: Cotton acreage on the Atlantic Coastal Plan in Georgia, more than doubled in the 1990's. This has created a significant need for data to evaluate performance of agrichemical runoff models for land in cotton production. This study focused on 3 chemicals widely used for pre-harvest defoliation. There are no published data available on how much of these products may be washed from treated fields. Simulated rainfall was applied to a series of small cotton plots one-hour after spraying. Thus a worst-case scenario for chemical runoff was evaluated. Although worst-case, the scenario was not unrealistic for the region. Weather records, show that intense storms that produce runoff occur frequently in September when much of the cotton is harvested. The amount of defoliants which ran off ranged from 4 to 15 % of that applied. It was also noted that although erosion rates on plots in conservation tillage were lower, the loss of tribufos, which binds strongly yto soil, was not affected. This is because the sediment from these plots was found to have a higher capacity to bind the compound than sediment from plots under conventional tillage. We also observed that chemical binding capacity of sediments decreased during runoff duration in both tillage treatments. Predictive models of agrichemical runoff will be improved by taking this variation into account and data obtained on the defoliants will enhance of the quality of assessments of human and ecological risks presented by use of these compounds.
Technical Abstract: Implementation of Clean Water Act TMDL provisions and tolerance reassessment under the Food Quality Protection Act has renewed interest in refining agrichemical runoff models. Cotton defoliant runoff under simulated rainfall was examined on 6-m2 plots in a cotton field in South Central Georgia. Dimethipin, thidiazuron and tribufos were applied to 3 plots each in portions of the field under strip and conventional tillage. One hour later, 50-mm of rainfall was applied in one hour. Long-term precipitation records indicate that the timing of rainfall application represented an extreme "worst-case" although not an unrealistic scenario for the region. Runoff was collected at 5-min intervals for the entire hour and tested for dissolved and sediment-bound forms of dimethipin, thidiazuron, and tribufos. Fractions of the chemicals applied that were lost in runoff were thidiazuron and tribufos 14-17 % and dimethipin 2-6 %. Timing of rainfall relative to defoliant application indicates that this i an upper-bound estimate of runoff losses. There was a pronounced "first- flush" effect with 30-60 % of the mass of each compound lost, detected in the first 5-minute sample while sample volume represented <3 % of the total hourly runoff. Soil losses on strip-till plots were less, but tribufos losses were not affected. Higher sediment-water partition coefficients (kd) observed in strip-till samples provided an explanation. In both tillage treatments tribufos-kd's were also observed to decrease with runoff duration. Predictive models of agrichemical runoff could be improved by taking this variation into account.