Variation of effect endpoint parameters in a terrestrial model ecosystem

Model ecosystems are widely used to assess the fate and the effects of chem icals in the environment, because they yield reproducible data on higher le vels of biological organization (populations, communities, ecosystems) at m oderate cost. Still, there is a need for the development of complex terrest rial test systems which allow for the statistical evaluation of results, Th e variation ('normal range') of possible effect endpoints in such systems i s of particular interest when comparing the ecotoxicological effects of che micals. However, experiments designed to assess the variation of endpoint p arameters are rarely performed, because they obviously do not yield data on the main topic, the effects of chemicals. In this study we present a laboratory terrestrial model ecosystem (TME) of 701 volume in which the most relevant components of a terrestrial ecosystem are represented by organisms easily kept in the laboratory. In the TME the re are four species of grass, three different arthropods and soil-microorga nisms, representing the important trophic levels of a terrestrial ecosystem . For these test organisms simple tests for sublethal parameters can be per formed both in single species tests and in the TME. In our experiment, the variation of possible effect endpoints from six untr eated systems was determined. The coefficient of variation (CV) of all endp oints studied here was between 8% and 77% and the CV of those endpoints cho sen for future studies was 8-48%. We believe that with increasing experienc e these values can even be lowered and that effects of 10-20% compared to c ontrol systems can be detected in the TME.