Modelling pesticide concentrations in Japanese paddy fields using the RICEWQ model

Abstract

Increasing concerns about plant protection products in surface waters have highlighted the importance of pesticide monitoring and modelling, both nowadays integral components of the pesticide registration process. The Rice Water Quality (RICEWQ) model predicts the fate and transport of pesticides under various paddy environmental conditions. The model has been used widely for regulatory purpose in the U.S., while its use in Europe has been limited to regulatory submissions, despite multiple recommendations to adopt RICEWQ as a high tier assessment model for regulatory purposes. The applicability of the RICEWQ model under Japanese agricultural conditions remains unexplored. This study leverages field experimental data to evaluate the RICEWQ model's capability to simulate daily concentrations of two herbicides, mefenacet and pretilachlor, in paddy water and paddy soil in Japan.The RICEWQ model provided a reasonable level of performance for predicting the fate and transport of the two herbicides by ensuring that the simulated daily water balance corresponded with field observations and with minor pesticide-specific calibration. To further improve the performance of the simulations, we implemented a straightforward calibration framework. Calibrating the RICEWQ model improved the accuracy of all simulations; for both herbicides and in both paddy water and paddy soil compartments, the lowest Nash-Sutcliffe efficiency achieved was 0.725, demonstrating excellent performance. However, the RICEWQ model consistently underestimated the peak concentrations of herbicides in paddy water within the first three days of simulation. Simulation results suggested that approximately 50 % of the total applied herbicide was lost from the paddy system. Increasing the duration of the water holding period after pesticide application or increasing the storage capacity of the rice field via excess water storage depth management has the potential to reduce greatly herbicide loss to below 10 % of the total applied herbicide.