RESUMEN
No-till and mulching are typical management operations in conservation agriculture (CA). To model pesticide degradation and leaching under a CA scenario, as compared to a conventional-tillage scenario (CT), the mulch module of the agro-hydrological model Daisy was extended. A Daisy soil column was parameterized with measurements of topsoil, mulch, and a realistic subsoil, and tested against published experimental data of pesticide fate in laboratory soil columns covered by mulch. Uncertainty and sensitivity analyses of the new Daisy version were conducted for a series of weather, soil, pesticide, and mulch parameters, using 4939 Monte Carlo simulations under each scenario. Results showed that there was no systematic difference in pesticide leaching from the topsoil (to the subsoil and directly to drains via drain-connected biopores) between CA and CT, but pesticide degradation and sorption were significantly different; degradation in the mulch and uppermost soil surface layer (0-3.5 cm) was larger in CA while degradation was larger in CT when considering the whole topsoil (0-30 cm). This difference for the whole topsoil could be explained by pesticide interception in CA in the part of the mulch not in direct contact with the soil where degradation is assumed not to occur. The sensitivity analysis highlighted non-influential parameters and seven parameters out of twenty-five to be better estimated to improve the accuracy of the predictions.
RESUMEN
Knowledge of the movement of herbicides and soil particles to sub-surface tile drainage may help to predict chemical leaching to surface waters and deeper groundwater systems. The movement of pendimethalin (2 years), ioxynil (1 year) and soil particles (3 years) to two tile drains was investigated on a sandy loam soil under natural weather conditions. Herbicide and particle concentrations in the drain water showed a very dynamic pattern. The largest herbicide concentrations were detected during the first tile drain flow events after application. Very little herbicide was lost with drain water later than 2 months after application. The turbidity, reflecting concentrations of soil particles, correlated positively and strongly with the pendimethalin concentration and negatively with the rate of drain water discharge, whereas it was uncorrelated with the ioxynil concentration. Peak turbidity values occurred during or shortly after rainfall events, either in break of frost situations, or on unfrozen soil coinciding with the occurrence of peak moisture contents in the topsoil well (3-7%) above field capacity. On average, 0.0013% of the applied pendimethalin and 0.0015% of the applied ioxynil were lost with drain water. The results suggest that preferential flow promotes the movement of all three substances to the tile drains but indicate somewhat different transport mechanisms for the two herbicides.