Pesticide transport with runoff from turf: observations compared with TurfPQ model simulations.

Pesticides applied to turf grass have been detected in surface waters raising concerns of their effect on water quality and interest in their source, hydrological transport and use of models to predict transport. TurfPQ, a pesticide runoff model for turf grass, predicts pesticide transport but has not been rigorously validated for larger storms. The objective of this study was to determine TurfPQ's ability to accurately predict the transport of pesticides with runoff following more intense precipitation. The study was conducted with creeping bentgrass [Agrostis palustris Huds.] turf managed as a golf course fairway. A pesticide mixture containing dicamba, 2,4-D, MCPP, flutolanil, and chlorpyrifos was applied to six adjacent 24.4 by 6.1 m plots. Controlled rainfall simulations were conducted using a rainfall simulator designed to deliver water droplets similar to natural rain. Runoff flow rates and volume were measured and water samples were collected for analysis of pesticide concentrations. Six simulations yielded 13 events with which to test TurfPQ. Measured mean percentage of applied pesticide recovered in the runoff for dicamba, 2,4-D, MCPP, flutolanil, and chlorpyrifos was 24.6, 20.7, 14.9, 5.9, and 0.8%, respectively. The predicted mean values produced by TurfPQ were 13.7, 15.6, 15.5, 2.5, and 0.2%, respectively. The model produced correlations of r=0.56 and 0.64 for curve number hydrology and measured hydrology, respectively. Comparisons of the model estimates with our field observations indicate that TurfPQ under predicted pesticide runoff during 69.5+/-11.4 mm, 1.9+/-0.2 h, simulated storms.