Modelling the fate of pesticides in paddy rice-fish pond farming systems in northern Vietnam

Abstract

In Vietnam, paddy rice fields have been identified as a major non-point source of pesticide pollution of surface- and groundwater which is often directly used for domestic purposes. One strategy to assess the risk of pesticide pollution is to use process-based models. Here, we present a new model developed for simulating short-term pesticide dynamics in combined paddy rice field-fish pond farming systems. The model was calibrated using the Gauss-Marquardt-Levenberg algorithm and validated against measured pesticide concentrations of a paddy field-fish pond system typical for northern Vietnam. In the calibration period, model efficiencies were 0.82 for dimethoate and 0.87 for fenitrothion. In the validation period, modelling efficiencies slightly decreased to 0.42 and 0.76 for dimethoate and fenitrothion, respectively. Scenario simulations revealed that a field closure period of 1 day after pesticide application considerably reduces the risk of pond and surface water pollution. These results indicate that the proposed model is an effective tool to assess and evaluate management strategies, such as extended field closure periods, aiming to reduce the loss of pesticides from paddy fields.