Water runoff and pesticide transport from a golf course fairway: Observations vs. Opus model simulations

Frequent pesticide applications to gulf courses causes concern that surface water may become contaminated, We hypothesized that runoff potential of th ese pesticides could be predicted by the recently developed Opus model. We conducted a 3-yr field study measuring surface runoff of water and dimethyl amine salts of 2,4-D [(2,4-dichlorophenoxy) acetic acid], dicamba (3,6-dich loro-2-methylphenoxy-benzoic acid), and mecoprop [(+/-)-2-(4-chloro-2-methy lphenoxy)-propanoic acid]. Twelve 7.4 m by 3.7 m plots of 'Tifway 419' berm udagrass (Cynodon dactylon (L.) Pers. x C. transvaalensis Burtt Davy) were managed as a golf course fairway. Simulated rainfall was applied at an aver age intensity of 29 mm h(-1) 1 d before and 1, 2, 4, and 8 d after pesticid e application for 0.92, 1.75, 1.75, 0.92, and 0.92 h, respectively. Average annual runoff loss was 9.13, 15.41, and 10.82% of applied 2,4-D, dicamba, and mecoprop, respectively. Both mass and concentration of pesticide runoff decreased rapidly, with the first posttreatment event runoff averaging 74. 5, 71.7, and 73.0% of the total runoff of 2,4-D, dicamba, and mecoprop, res pectively. The Opus model adequately simulated runoff [R-2 = 0.897 and norm alized root mean square error (NRMSE) = 24.6%], The 2,4-D in runoff was bet ter simulated by complete-kinetic sorption (R-2 = 0.876, NRMSE = 60.2%) tha n by equilibrium sorption (R-2 = 0.848, NRMSE = 68.2%). Opus did not accura tely simulate 2,4-D over all runoff events, but simulated 2,4-D in the firs t posttreatment runoff within a factor of 2 of those measured.