A site-specific farm-scale GIS approach for reducing groundwater contamination by pesticides

It has been proposed to vary pesticide applications by patterns in surface organic C to reduce the potential for contamination of groundwater. To evaluate the feasibility of this 'precision farming' approach, data for carbofuran concentrations from 57 locations sampled to a depth of 180 cm were fit to the convective-dispersive equation. Fitted values for pore water velocity (v) ranged from 0.17 to 1.92 cm d^-1, with a mean of 0.68 cm d^{- 1}. Values for dispersion (D) ranged from 0.29 to 13.35 cm² d^-1, with a mean of 3.49 cm² d^-1. Values for the retardation factor (R) ranged from 1.28 to 6.83, with a mean of 2.57. With this data, risks of pesticide leaching were estimated at each location using the attenuation factor (AF) model, and a dispersion based leached factor (LF) model. Using the AF model gave two locations with a very high pesticide leaching risk, 32 with a high pesticide leaching risk, 15 with a medium risk, 6 with a low risk, and 2 with no risk. Using the LF model, 6 had a high risk, 13 had a medium risk, 18 had a low risk, and 20 had no risk. Pesticide leaching risks were not correlated with any measured surface soil properties. Much of the variability in leaching risk was because of velocity variations, so it would be incorrect to assume that surface organic C content controls the leaching risk.