A long-term (>10 months) leaching experiment was conducted with a large clay soil column and a rain simulator to study unsaturated transport of the nematicide aldicarb and the herbicide simazine in a cracked clay soil. Water retention and soil conductivity were derived from experimental outflow data and deterministic parameter estimation techniques. Under conventional application rates and realistic rain events, aldicarb's aerobic metabolites were found in very high concentrations, and did not meet the normal EC levels for water during the entire measuring period. A mass balance for aldicarb showed that 0.35% of the initial dose was leached. However, when the two isosteric metabolites aldicarb-sulfoxide and aldicarb-sulfone were included in the mass balance, this percentage increased dramatically to 19.7%. Simazine was found in relatively low concentrations of 0.05-0.6 g/l, and only 0.11% of the initial dose was leached over 280 days. The absence of a `breakthrough behaviour' (peak exposure) implies long term delivery (chronic exposure) of the compound from the soil. The predictive performances of the widely used pesticide leaching models VARLEACH, LEACHP, MACRO, PESTLA and SIMULAT, which differ in their basic concepts for calculating water and solute transport and pesticide behaviour, were compared. This ring test revealed that none of the models were able to describe both water percolation and pesticide leaching to a completely satisfying degree. Moreover, there is little agreement on maximum pesticide concentrations and the time period in which these occur. This conclusion seriously limits the possibilities of model application and conducting reliable risk assessments for pesticides which are applied on the studied, or similar type of clay soils.
- plant protection
- pesticide residues
- hydraulic conductivity
- clay soils
Vink, J. P. M., Gottesbüren, B., Diekkrüger, B., & van der Zee, S. E. A. T. M. (1997). Simulation and model comparison of unsaturated movement of pesticides from a large clay lysimeter. Ecological Modelling, 105, 113-127. https://doi.org/10.1016/S0304-3800(97)00147-6