Fate of Pesticides in Surface Waters, Laboraty and Field Experiments

Fate of Pesticides in Surface Waters, Laboraty and Field Experiments

Summary

The work reported here was performed as part of the project ”Development of an operational and validated model for pesticide transport and fate in surface water” which is also termed Subproject 5. Subproject 5 is a subproject of the main project: ”Model based tool for evaluation of exposure and effects of pesticides in surface water”. The overall goal of the main project was to provide the Danish EPA with a tool that, on the basis of pesticide property data, could be used to evaluate the fate of new pesticides in surface water, within certain scenarios, in relation to registration and regulation. In effect, the heart of the tool is a full catchment fate and transport model for two existing catchments, the Odderbæk and the Lillebæk catchment, that each contains a small stream. Subproject 5 has been concerned with the transport and fate in streams and ponds in these catchments.

This report contains the results of laboratory and field experiments made in Subproject 5. The objectives of the laboratory and field experiments were to gather information on processes critical for fate and transport of pesticides in streams and ponds in order to decide which processes to include in the model, to evaluate the applicability of currently required registration data for modelling and to gather parameter values to be used in the evaluation of the models.

The three pesticides pendimethalin, ioxynil and bentazone were used in the laboratory experiments while the field experiments also included fenpropimorph and glyphosate. The pesticides were chosen to get a wide variety in physico-chemical properties.

The laboratory experiments were set up to investigate sorption and degradation processes. As the streams to be modelled are very small and rather short, the residence time of a pesticide entering the streams will be short. Therefore, both sorption equilibrium and sorption kinetics were investigated. The effects of temperature, pesticide concentration level and sediment properties on the sorption of the selected pesticides were investigated to determine how to set up the model and to get an idea of how well pesticide property data delivered by the registration applicant could be used in the model-based tool.

The sorption processes were investigated in batch experiments where sediment from the catchment streams, the pond used in the field experiments and from a number of lakes were shaken with water and radiolabelled pesticide for different periods of time. Concentrations were measured indirectly by scintillation counting.

The sorption experiments showed that the pesticides sorbed to the sediments according to their hydrophobicity and the sediment organic matter content. Furthermore, as the Koc values measured for stream, pond and lake sediments were in accordance with data for soil as found in the registration material, it seems reasonable to use soil-derived Koc values for the modelling of these types of pesticides. Pendimethalin sorbed rather strongly while ioxynil and bentazon sorbed much less.The sorption of the strongly sorbing pesticide pendimethalin showed considerable non linearity as sorption was higher at low concentrations than at high, but a Freundlich isotherm expression did not describe data well.

There was no difference between sorption at 4° C and 20° C for the three pesticides and the influence on sorption of temperature variations, which can be expected in the surface water of the catchments is thus considered negligible.

Sorption was fast in all cases. Waterphase concentrations very close to equilibrium were reached within a few hours. However, the rate of sorption is probably not fast enough to be ignored in the stream model. This will be shown by model runs. Variations in initial pesticide concentration did not affect the rate of sorption for a sediment with low organic matter content.

Aerobic degradation processes in stream sediment-surface water suspensions and in surface water were investigated in batch experiments with radiolabelled pesticides. Bulk samples were taken from the water and suspension at regular intervals for approx. 100 days and counted.

The degradation experiments showed that the pesticides degraded quite differently. Seemingly, bentazone did not degrade at all in the experiments, ioxynil degraded slowly but steadily and pendimethalin initially disappeared quite fast but, after a period, no more reduction in activity could be measured. This observation on pendimethalin degradation behaviour could be attributed to a number of reasons of which incorporation into biomass seems to be the most plausible.

The experiments showed that, for the weakly sorbing pesticide ioxynil, degradation rates were higher in the sediment-water suspensions than in water alone, which may be due to the higher biomass concentration in the suspensions.

For the strongly sorbing pesticide pendimethalin, the degradation rates in sediment suspensions and surface water were similar but the level of relative disappearance reached was higher for surface water alone than for surface water-sediment suspensions. This may be attributed to strong sorption of the pesticide to the solids of the suspension, which reduces bioavailability.

A “fate” model including first order sorption, desorption and degradation was successfully fitted to the experimental data and this “fate” model is incorporated in the fate and transport model of the model-based tool. The fitting of the fate model produced a large number of sorption and degradation rates that will be used in the evaluation of the full fate and transport model.

The field experiments were conducted in order to investigate the dissipation time from water for pesticides having different physico-chemical properties and the vertical mixing in more or less stagnant ponds with and without macrophytes, and to get data for evaluation of the effective diffusion coefficient.

The field experiments were conducted in artificial ponds by spraying the pond water surfaces with formulations of the five different pesticides and following the development in concentrations in the water body, at different depths, and in the sediment. The ponds were well developed with both flora and fauna and thus resembled natural shallow ponds. The pesticide content of water and sediment samples was determined by chemical analysis. In the case of glyphosate, analysis for AMPA, the major degradation product, was also performed.

The two hydrophobic pesticides pendimethalin and fenpropimorph had dissipation half times from the water phase of around one to two days and the dissipation rate for glyphosate was slightly lower. Ioxynil disappeared from the water phase within 20 days while bentazone, being more persistent, could still be measured after 130 days.

There was a large difference between water phase dissipation rates measured in two different years, which demonstrates the importance of variation in conditions. One year, a heavy rain shower in the days before spraying made it necessary to pump water from ponds, which might have caused resuspension of sediment. That year, dissipation rates were markedly higher than the year before when no similar rain event had occurred.

The presence of macrophytes in the ponds did not influence dissipation rates of fenpropimorph and pendimethalin very much and dissipation rates from ponds with macrophytes and ponds without macrophytes were quite similar. However, a statistical analysis showed that the presence of macrophytes initially causes a higher concentration in the water phase whereas, after one day, it causes a lower concentration in the water phase when compared to macrophyte-free ponds. This may be explained by the combined effect of the macrophytes on turbulence and sorption. Macrophytes hinder turbulence and thus initially reduce the movement of pesticides from the surface to the sediment, which results in a higher water concentration. However, after a while the water body is eventually mixed in spite of the macrophytes, and now the macrophytes act as a sorbing compartment that removes pesticide from the water phase and thus reduces concentration.

The experiments showed that two days after spraying, a vertical concentration gradient could still be measured in ponds sprayed with pendimethalin and fenpropimorph. For glyphosate, a gradient was only discernible on the first day after spraying.

The concentrations of pendimethalin and fenpropimorph in sediment 2, 15 and 31 days after spraying were higher in ponds without macrophytes than in ponds with macrophytes. The reason again seems to be the reduction in vertical mixing and thereby reduced transport from surface to bottom, which the macrophytes cause, and the sorption of pesticide to macrophytes.