Simulated Regional Variations in Pesticide Runoff

Abstract

AbstractMathematical models for daily weather, hydrology, and soil chemistry were combined in a general approach for estimating pesticide runoff. Runoff losses of atrazine [2‐chloro‐4(ethylamino)‐6‐(isopropylamino)‐s‐triazine] and carbofuran (2,3‐dihydro‐2,2‐dimethyl‐7‐benzofuranyl methylcarbamate) were simulated for hypothetical 9‐ha corn (Zea mays L.) fields in New York, Georgia, Iowa, Texas, and Kentucky. Two representative soils were considered: a well‐drained sandy loam and a poorly‐drained silty clay loam exemplified by a Typic Dystrochrept and Typic Albaqualf, respectively. The two pesticides, five locations, and two soils produced a total of 20 cases. Each case was simulated for 100 yr by a Monte Carlo process to produce 100 samples of annual pesticide runoff. Mean annual pesticide losses ranged from 0.7 to 6.3% of application for atrazine and from 0.8 to 11.4% for carbofuran. Statistical analyses of the samples indicated that soil type affects pesticide runoff, with the silty‐clay loam producing significantly greater losses than the sandy loam. Regional similarities and differences were also identified. New York pesticide losses were significantly lower than those from the other four sites. Conversely pesticide runoff values from Georgia and Texas were statistically similar. Variations in pesticide runoff are largely explained by seasonal runoff volume. Regressions on mean runoff volume during the month of pesticide application explained over 89% of the observed variation in mean annual pesticide in runoff.