Use of Microcosms for Evaluation of Toxic Substances

Microcosms are designed to help understand certain ecosystem processes, rather than to replicate functional ecosystems. The rationale for their use is that they provide information (e.g. on the behaviour of toxic substances) at the process level. The microcosm is not an undisturbed natural system, but retention of critical subsystems enables it to be used as a simulation of nature. The major problems which have been encountered in the use of microcosms concern replicability (precision); in general, the larger or more complex the system, the greater its value for the maintenance of reality, but the poorer its reproducibility. Several points should be considered when applying the microcosm as a tool in toxicology: (1) a hypothesis should be tested; (2) the experiment must be realistic; although (3) pragmatically the design is a simulation,not a reproduction of an ecosystem; (4) testing should be at the process level; and (5) reproducibility should be demonstrated. Examples are presented of complex microcosms which have given reproducible results. It is recommended that microcosms are useful research tools when used in conjunction with other approaches (bench top experiments and field observations).