ABSTRACT
Semi-field tests cover a large number of methodological approaches which combine the advantages of laboratory testing and field tests in one form or another. One main problem involved with semi-field approaches is that the results of pesticide evaluations are difficult to generalize if the test design was sharply defined by local factors. Another — and even more frequent — problem to date has arisen if a test was, for example, especially designed for site evaluation to determine whether or to what extent the site should be protected. Model ecosystem segments or microcosms are understood to mean systems which can be controlled and reproduced (in the laboratory, greenhouse or even in climate chambers) to simulate the processes and interactions of the chosen segment of the environment.
