Modeling pesticide leaching and dissipation in a mediterranean littoral greenhouse

The one-dimensional chromatographic flow model PEARL was used to simulate the movement of the insecticide imidacloprid and the fungicide procymidone through a greenhouse soil. The model was parametrized using measured and literature values of soil hydrological parameters. Soil water movement and soil temperature were reasonably well described by the model. The ability of PEARL to simulate the fate of imidacloprid and procymidone following four applications of each compound was evaluated against greenhouse data. Simulated imidacloprid residues in the 0-10 cm layer were in good agreement with measured data. Below 10 cm, the model overestimated imidacloprid remaining following the spray applications, whereas simulated residues following the chemigation applications were in reasonable agreement with measured data. Simulated residues of procymidone in the 0-10 cm layer were in general agreement with measured values. In the 10-20 cm layer, peaks in simulated concentrations occurred later than observed in the greenhouse. Predictions of procymidone residues below 20 cm were poor and underestimated compared to the measured data. For both pesticides, discrepancies between modeled and measured data in the 10-20 cm layer were attributed to the drip irrigation method used in the greenhouse. The model was unable to satisfactorily predict pesticide movement from the soil surface by irrigation water unless the scenario was modified to reflect the localized pattern of water application. Scenario analysis indicated that air boundary layer thickness is a key parameter for readily volatilized pesticides such as procymidone. This is of particular relevance to the greenhouse environment, where the boundary layer thickness may be greater than that in outdoor conditions.