Introduction to Microcosm and Mesocosm Tests

There is a wide spectrum of these microcosm and mesocosm higher-tier test systems available, from laboratory microcosms to outdoor mesocosms, enclosures, and artificial streams (ECETOC 1997). Fish are usually not included, and this is a problem if single-species tests suggest that fish are more sensitive than algae or invertebrates (Girling et al. 2000). 

Studies for pesticide risk assessment focus on the situation in a water body near the field edge, with the peak of exposure soon after application due to drift, runoff, or drainage. In most cases, endpoints are related to the initial concentration of the test item (which can encompass multiple applications). In contrast to this, the focus of EQS derivation is mostly on protection against effects of long-term exposure. 

Endpoints available to date are most often presented as NOECs and rarely ECx values. Compared to laboratory tests, the number of replicates is usually lower and the variability between replicates higher (due to complexity of the systems and study durations of several weeks or months). Thus, the statistical power might be low, especially for rarer taxa. Due to these limitations, assessment of effects should not be restricted to statistical NOECs alone but also based on inspection of plots of population dynamics. Multivariate statistics are generally used to test for effects on community structure. 

For pesticide risk assessment, recovery is taken into account to derive an NOEAEC (no-observed ecologically adverse effect concentration). A maximum of 8 weeks is often taken as an acceptable recovery period, but the life cycle of the affected species should also be taken into consideration (SANCO 2002). 

---