Umsetzung und Anwendungen
(Englisch)
|
Summary and Conclusions
Within the scope of the OECD pesticide risk reduction program a project for the development of pesticide aquatic risk indicators was initiated in 1998. As a result, three indicators (REXTOX, ADSCOR and SYSCOR) were developed to estimate relative risk of agricultural pesticide use to surface water organisms. They were designed as political tools, for national authorities to follow progress of risk reduction measures or plan pesticide management.
The indicators combine pesticide properties, pesticide use data such as applied dose rate, application characteristics such as method of application, and some environmental parameters. In contrast to various existing approaches which calculate hazard indices based on worst case assumptions, the OECD aquatic risk indicators calculate risk indices based on actual use data. In all three indicators relative risk values are estimated by calculating the «exposure:toxicity» ratio. While all indicators include toxicity to the same target organisms (algae, daphnia, and fish), they differ in the approach of exposure estimation (Chapter 2). REXTOX uses a mechanistic approach and is the only indicator considering several field site properties. ADSCOR and SYSCOR are scoring indicators, assigning scores to variables assumed to contribute to aquatic risk.
In 2000, the OECD aquatic risk indicator pilot project started as follow up to the indicator development project. Its scope was to allow several countries to test the indicators with their own input data, to assess the indicators' usefulness, validity, and acceptability as policy tools. Apart from Switzerland, Denmark, France, Germany, Japan, Norway, and USA participated in this project.
Input data
Compilation of reliable input data for the indicators (Chapter 3) turned out to be time consuming. In general properties of active ingredients were derived from handbooks or registration dossiers. For most pesticides physico-chemical data such as log Kow and solubility were readily available. In contrast, for degradation rates in natural water and long term toxicities significant data gaps or data variability of several orders of magnitude were encountered. Swiss use data were only available for the years 1997 and 1998 for the catchment areas of three lakes, namely Greifensee, Baldeggersee, and Murtensee. The limited availability of use data did not allow the calculation of time trends or extrapolation to a national level. In general, the quality of input data, i.e. pesticide use data, physico-chemical data of pesticides, and toxicity data of pesticides has to be improved.
Validation
The Swiss contribution to the indicator validation, included three parts: Sensitivity analysis, exemplary and scenario calculations, and comparison of indicator results with monitoring data. Only two of the three indicators could be thoroughly tested, because the third (SYSCOR) was still under development during the final phase of the project. For the two other indicators it was shown, that apart from toxicity the indicators primarily are driven by the scaling variables «applied amount of pesticides» (REXTOX) or «area treated» (ADSCOR).
It became obvious that revision and adaptation to the specific Swiss situation would be necessary for both, ADSCOR and REXTOX. Thereby, most important in REXTOX is the need for recalibration of the relative weight of the two considered input pathways of pesticides to surface waters «spray drift» and «runoff», as the former is clearly overestimated. In ADSCOR, the combination of different scales for variables such as scores and real values has to be reconsidered carefully.
It was found that REXTOX may be useful as add-on tool for priority setting of monitoring programs as the indicator reflects the relative portion of pesticides found in surface water, in general, quite well.
Use of indicators
The indicator results can be presented in different ways, in order to identify causes for relative risk changes, regional differences or «risky» crops or pesticides, depending on the questions to be addressed.
Apart from above mentioned need for revision, REXTOX and ADSCOR are seen as useful tools to interpret actual pesticide use data in combination with pesticide properties with respect to relative risks to aquatic organisms. Nevertheless, it should be emphasised, that indicator results may never be used as such, but need to be analysed carefully and the interpretation has to be committed to experts. Thereby the limitations of indicators concerning their meaningfulness and completeness have to be well considered and communicated. Indicators should never be used as single basis for decision making and the use of several indicators instead of only one seems to be appropriate.
At the time of writing this report it was not yet decided by the responsible Swiss authorities whether one or several of the three OECD indicators will be considered in the future. However, prior to further development or implementation of risk indicators in Switzerland, it is necessary that specific requirements, such as questions to be addressed, variables to be considered, intended users and target audience, and available resources are defined.
|
