Collembola in ecotoxicology—Any news or just boring routine?

• Soil properties, species biology and test conditions affect the outcome of a test. • The effort for the OECD reproduction test can be substantially reduced. • The miniaturised test presented here saves time, animals and chemicals. • The outcome of choice tests is affected by both intra- and interspecific interactions. Despite the uncontested significance of soils for human nutrition and drinking water quality, the majority of ecotoxicological testing is confined to aquatic test systems. Among the standardised tests for soils, the reproduction test with the springtail Folsomia candida is among the most widely used ones. First steps towards its standardisation were undertaken in the late 1980s. Here we review major advances that have been made since then, with respect to mechanistic, pragmatic and ecological aspects. Specifically we address the ecological relevance of any modifications of the standardised tests. We introduce a miniaturised version of the reproduction test which allows reducing the amount of soil per test unit to one third and the number of synchronised individuals to 40% as compared to the standard test. In addition, we developed an assay using Collembola eggs instead of synchronised adults. First results of a three-species test indicate that the presence of other species may affect choice behaviour. We point out a potential biased view of existing ecotoxicological data with Collembola due to the fact that most results refer to metal contamination. Finally, recommendations for future research are given, with special reference to avoidance and microcosm tests involving Collembola.