Julien Boulange

Simulating the fate and transport of nursery-box-applied pesticide in rice paddy fields

BACKGROUND The Pesticide Concentration in a Paddy Field model (PCPF-1) was modified by adding a root zone compartment to simulate nursery-box-applied (NB-applied) pesticide. The PCPF-NB model was validated for predicting the concentrations of NB-applied fipronil and imidacloprid in rice paddy fields using two treatment methods: before transplanting (BT) and at sowing (AS). Uncertainty and sensitivity analyses were used to evaluate the robustness of the concentrations predicted by the model. RESULTS The hourly predicted concentrations of imidacloprid and fipronil were accurate in both paddy water and 1 cm deep paddy soil. The coefficient of determination and Nash-Sutcliffe model efficiency were greater than 0.87 and 0.60 respectively. The 95th percentiles of the predicted concentrations of fipronil and imidacloprid indicated that the influence of input uncertainty was minor in paddy water but important in paddy soil. The pesticide deposition rate and the desorption rate from the root zone were identified to be the major contributors to the variation in the predicted concentrations in paddy water and soil. CONCLUSION The PCPF-NB model was validated for predicting the fate and transport of NB-applied fipronil and imidacloprid using the BT and AS treatment methods.

Improvement and application of the PCPF-1@SWAT2012 model for predicting pesticide transport: a case study of the Sakura River watershed: Improvement and application of the PCPF-1@SWAT2012 model

BACKGROUND The Soil and Water Assessment Tool combined with Pesticide Concentration in Paddy Field (PCPF-1@SWAT) model was previously developed to simulate the fate and transport of rice pesticides in watersheds. However, the current model is deficient in characterizing the rice paddy area and is incompatible with the ArcSWAT2012 program. In this study, we modified the original PCPF-1@SWAT model to develop a new PCPF-1@SWAT2012 model to address the deficiency in the rice paddy area and utilizing the ArcSWAT2012 program. Next, the new model was applied to the Sakura River watershed, Ibaraki, Japan in order to simulate the transport of four herbicides: mefenacet, pretilachlor, bensulfuron-methyl and imazosulfuron. RESULTS The results showed that the water flow rate simulated by PCPF1@SWAT2012 was similar with the observed data. The calculated Nash-Sutcliffe efficiency coefficient (NSE) (0.73) and percent bias (PBIAS) (-20.38) suggested satisfactory performance of the model. In addition, the concentrations of herbicides simulated by the PCPF-1@SWAT2012 model were in good agreement with the observed data. The statistical indices NSE and root mean square error (RMSE) estimated for mefenacet (0.69 and 0.18, respectively), pretilachlor (0.86 and 0.18, respectively), bensulfuronmethyl (0.46 and 0.21, respectively) and imazosulfuron (0.64 and 0.28, respectively) indicated satisfactory predictions. CONCLUSION The PCPF-1@SWAT2012 model is capable of simulating well the water flow rate and transport of herbicides in this watershed, comprising different land use types, including a rice paddy area.

PCPF‐M model for simulating the fate and transport of pesticides and their metabolites in rice paddy field

BACKGROUND The PCPF-1 model was improved for forecasting the fate and transport of metabolites in addition to parent compounds in rice paddies. In the new PCPF-M model, metabolites are generated from the dissipation of pesticide applied in rice paddies through hydrolysis, photolysis and biological degradations. The methodology to parameterize the model was illustrated using two scenarios for which uncertainty and sensitivity analyses were also conducted. RESULTS In a batch degradation experiment, the hourly forecasted concentrations of fipronil and its metabolites in paddy water were very accurate. In a field-scale experiment, the hourly forecasted concentrations of fipronil in paddy water and paddy soil were accurate while the corresponding daily forecasted concentrations of metabolites were adequate. The major contributors to the variation of the forecasted metabolite concentrations in paddy water and paddy soil were the formation fractions of the metabolites. The influence of uncertainty included in input parameters on the forecasted metabolite concentration was high during the peak concentration of metabolite in paddy water. In contrast, in paddy soil, the metabolite concentrations forecasted several days after the initial pesticide application were sensitive to the uncertainty incorporated in the input parameters. CONCLUSION The PCPF-M model simultaneously forecasts the concentrations of a parent pesticide and up to three metabolites. The model was validated using fipronil and two of its metabolites in paddy water and paddy soil. The model can be used in the early stage of the pesticide registration process and in risk assessment analysis for the evaluation of pesticide exposure.

Aspect of the degradation and adsorption kinetics of atrazine and metolachlor in andisol soil

Summary The degradation kinetics and sorption characteristics of atrazine and metolachlor in Japanese andisol soil were evaluated using laboratory incubation of soil samples. The water content of the soil was set to field capacity while three different temperatures (5, 25 and 35°C) were considered for the experiment. First order model fitted the degradation kinetics of both herbicides under the investigated temperature range with half-lives ranging from 19.2 to 46.9 days for atrazine and from 23.4 to 66.9 days for metolachlor, respectively. The activation energies (Ea) of atrazine and metolachlor calculated using Arhenius equation were 21.47 and 23.91 kJ mol−1, respectively. The soil sorption study was conducted using the batch equilibrium process. The adsorption behaviors of atrazine and metolachlor were investigated using linear, Freundlich and Langmuir isotherms although the linear and Freundlich isotherms gave relatively high correlation coefficient (R2) and very low standard error of estimate (SEE). The free energy (ΔG°) values were in the range −30.6 to −32.0 kJ/mol, and −32.1 to −41.5 kJ/mol for atrazine and metolachlor, respectively. Thermodynamic parameters indicated that the adsorption is spontaneous, endothermic accompanied by increase in entropy. The understanding of atrazine and metolachlor sorption processes is essential to determine the pesticide fate and availability in soil for pest control, biodegradation, runoff and leaching.

Development and validation of the SPEC model for simulating the fate and transport of pesticide applied to Japanese upland agricultural soil

A pesticide fate and transport model, SPEC, was developed for assessing Soil-PEC (Predicted Environmental Concentrations in agricultural soils) for pesticide residues in upland field environments. The SPEC model was validated for predicting the water content and concentrations of atrazine and metolachlor in 5-cm deep soil. Uncertainty and sensitivity analyses were used to evaluate the robustness of the models predictions. The predicted daily soil water contents were accurate regarding the number of observation points (n=269). The coefficient of determination (R 2) and Nash-Sutcliffe efficiency (NSE ) were equal to 0.38 and 0.22, respectively. The predicted daily concentrations of atrazine and metolachlor were also satisfactory since the R 2 and NSE statistics were greater than 0.91 and 0.76, respectively. The field capacity, the saturated water content of the soil and the Q 10 parameter were identified as major contributors to variation in predicted soil water content or/and herbicide concentrations.