ABSTRACT
Aquatic field tests have not been required in the German registration process thus far. However, some were submitted by several companies to support the registration of pesticides which were considered as very hazardous to aquatic systems. Thus, German authorities gathered some experience in interpreting results derived from different types of static mesocosms.
For future use in Germany the purpose of such systems is clear: like very similar regulations in other countries, the German Plant Protection Act requires the identification of possible effects of pesticide use on non-target systems and the probability of their occurrence. At present this means extrapolation from laboratory data (fate or effects) to real ecosystems. The use of mesocosms shortens this distance of extrapolation mainly by validating fate and bioavailability predictions and by observing effects at the community level, including compensation or recovery. Further, more experience with field tests will provide more general rules for hazard assessments.
To reach these objectives, mesocosms should be tailored to the substance of concern, although based on a general design. They are not likely to replace the single-species tests which are necessary to identify substance-specific problems. Also, field tests are unlikely to be required obligatorily for every compound. Their regulatory use should be restricted to pesticides with high risk to aquatic life (not necessarily only the extremely toxic ones but the persistent and/or frequently used ones as well). As unexpected effects cannot be ruled out, it should be considered to test representatives of every important compound group. More experience is required of effects on flowing systems.
As to experimental designs, the farm pond approach (combining effects testing and quantifying routes of entry in the same study) is not considered as very promising since its results are very site-specific. More flexibility in the use of the results and easier interpretation can both be achieved by having separate, more generalized models for exposure. However, the use-specific main route of entry should be modelled in the mesocosm study.
