Pesticide leaching to ground water at 1 m depth and pesticide persistence in the plow layer were calculated with a mathematical model for a sandy soil continuously cropped with maize (Zea mays L.) and exposed to weather conditions in a temperate climate. The pesticide was applied in spring. In the model, water flow was described by Darcy's law and water uptake by the crop was included. Dally averages of meteorological conditions (rainfall, evapotranspiration, soil temperature) were used as input. The model assumes first‐order transformation, equilibrium sorption (Freundlich equation), and passive plant uptake. Pesticide leaching and persistence were calculated as a function of pesticide sorption (characterized by the organic‐matter/water distribution coefficient, Kom) and of transformation rate. It was found that pesticide leaching is very sensitive to both Kom and the transformation rate: changing Kom or the transformation rate by a factor of 2 changes the fraction of the dose leached typically by about a factor of 10. Pesticide persistence in the plow layer was found to be sensitive to Kom at low transformation rates and sensitive to the transformation rate at high Kom values. Additional calculations showed that autumn application results in much higher leaching of nonsorbing pesticides with short half‐lives than spring application (difference of two orders of magnitude).

No abstract

The EFSA Scientific Committee addressed in this document the peculiarities related to the genotoxicity assessment of chemical mixtures. The EFSA Scientific Committee suggests that first a mixture should be chemically characterised as far as possible. Although the characterisation of mixtures is relevant also for other toxicity aspects, it is particularly significant for the assessment of genotoxicity. If a mixture contains one or more chemical substances that are individually assessed to be genotoxic in vivo via a relevant route of administration, the mixture raises concern for genotoxicity. If a fully chemically defined mixture does not contain genotoxic chemical substances, the mixture is of no concern with respect to genotoxicity. If a mixture contains a fraction of chemical substances that have not been chemically identified, experimental testing of the unidentified fraction should be considered as the first option or, if this is not feasible, testing of the whole mixture should be undertaken. If testing of these fraction(s) or of the whole mixture in an adequately performed set of in vitro assays provides clearly negative results, the mixture does not raise concern for genotoxicity. If in vitro testing provides one or more positive results, an in vivo follow‐up study should be considered. For negative results in the in vivo follow‐up test(s), the possible limitations of in vivo testing should be weighed in an uncertainty analysis before reaching a conclusion of no concern with respect to genotoxicity. For positive results in the in vivo follow‐up test(s), it can be concluded that the mixture does raise a concern about genotoxicity.

Assessment of risks to aquatic organisms is important in the registration procedures for pesticides in industrialised countries. This risk assessment consists of two parts: (i) assessment of effects to these organisms derived from ecotoxicological experiments (=effect assessment), and (ii) assessment of concentration levels in relevant environmental compartments resulting from pesticide application (=exposure assessment). Current procedures lack a clear conceptual basis for the interface between the effect and exposure assessments which may lead to a low overall scientific quality of the risk assessment. This interface is defined here as the type of concentration that gives the best correlation to ecotoxicological effects and is called the ecotoxicologically relevant concentration (ERC). Definition of this ERC allows the design of tiered effect and exposure assessments that can interact flexibly and efficiently. There are two distinctly different exposure estimates required for pesticide risk assessment: that related to exposure in ecotoxicological experiments and that related to exposure in the field. The same type of ERC should be used consistently for both types of exposure estimates. Decisions are made by comparing a regulatory acceptable concentration (=RAC) level or curve (i.e., endpoint of the effect assessment) with predicted environmental concentration (=PEC) levels or curves (endpoint of the exposure assessment). For decision making based on ecotoxicological experiments with time-variable concentrations a tiered approach is proposed that compares (i) in a first step single RAC and PEC levels based on conservative assumptions, (ii) in a second step graphically RAC and PEC curves (describing the time courses of the RAC and PEC), and (iii) in a third step time-weighted average RAC and PEC levels.

Summary: The adsorption of atrazine and its transformation products, desisopropylatrazine (2‐chloro‐4‐ethylamino‐6‐amino‐l,3,5‐triazine), desethyl‐atrazine (2‐chloro‐4‐amino‐6‐isopropylamino‐l,3,5‐triazine) and hydroxyatrazine (2‐hydroxy‐4‐ethylamino‐6‐isopropylamino‐l,3,5‐triazine) to four top‐soils was measured. Adsorption coefficients decreased in the order hydroxy atrazine, atrazine, desisopropylatrazine, and desethyl‐atrazine: the distribution coefficient between organic matter and water (KOM) ranged from 40 to 100 dm3 kg−1 for atrazine, from 30 to 60 dm3 kg−1 for desisopropylatrazine, from 20 to 50 dm3 kg−1 for desethylatrazine and from 100 to 590 dm3 kg−1 for hydroxy atrazine. Data are discussed in the context of earlier literature.

After calibration, PEARL could simulate well the observed rapid movement towards drains of two pesticides with contrasting sorption and degradation rate properties. The calibrated value for the fraction of the internal catchment domain was high (90%). This means that a large fraction of water entering the macropores infiltrates into the soil matrix, thus reducing the fraction of rapid flow.

The sensitivity of pesticide leaching to pesticide/soil properties and to meteorological conditions was assessed by calculations with an existing convection—dispersion model. The model assumes equilibrium sorption (Freundlich equation), first‐order transformation kinetics and passive plant uptake. The extent of pesticide leaching was characterized by the percentage of the dose leached below 1 m depth. The calculations were carried out for a humic sand soil cropped with maize and exposed to Dutch weather conditions. In general, the percentage leached was found to be very sensitive to the sorption coefficient, the Freundlich exponent (describing the curvature of the isotherm) and the transformation rate. The percentage leached was moderately sensitive to weather conditions (wet/dry years), long‐term sorption equilibration and the relationship between transformation rate and temperature. Sensitivity to the extent of plant uptake was only significant for pesticides with low sorption coefficients. Sensitivity to soil hydraulic properties was small. The effect of application in autumn instead of in spring was found to be very large for non‐sorbing pesticides with short half‐lives. The sensitivity to spatial variability in sorption coefficient and transformation rate was found to be substantial at low percentages leached.

Modelling is an economic way of assessing pesticide behaviour under ®eld conditions; it is cheaper and faster than ®eld experiments. Modelling attempts to generalize knowledge of pesticide ®eld behaviour through identi®cation of the most important pesticide/soil properties that can be measured in the laboratory. The technology to simulate volatilization of volatile pesticides that are incorporated or injected into the soil is well developed. However, modelling of volatilization rates from plant and soil surfaces before the ®rst signi®cant rainfall event after application is barely possible with current knowledge. The technology to simulate pesticide persistence in the plough layer is well developed; the PERSIST model has been tested at least 178 times, usually resulting in a slightly faster decline in the ®eld than was simulated. In general, available pesticide leaching models are reliable enough to assess the leaching of the bulk of the dose (leaching levels above 1%). The EU drinking water limit of 0.1 lg L )1 implies leaching of less than 0.1% of a dose of 1 kg ha )1 . At such a low leaching level, the validation status of the models is still low, mainly because preferential¯ow processes in both structured and unstructured soils and the factors controlling the transformation rate in subsoil are not well enough understood.