RESUMEN
Laboratory ecotoxicological tests are important tools for the management of environmental changes derived from anthropogenic activities. Folsomia candida is usually the model species used in some procedures. However, this species may not be sufficiently representative of the sensitivity of the other collembolan species. This study aimed to evaluate (i) the effects of soils naturally rich in potentially toxic elements (PTE) and soil characteristics on the reproduction and survival of different collembolan species, (ii) whether the habitat function of these soils is compromised, and (iii) to what extent F. candida is representative of the other collembolan species. For this, reproduction tests with six collembolan species were conducted in 14 different samples of soils. In general, collembolan reproduction was not completely inhibited in none of the natural tested soils. Even soils with high pollution load index values did not negatively affect collembolan reproduction for most of the species. In contrast, the lowest collembolan reproduction rates were found in a visually dense soil (lowest volume/weight ratio), highlighting that soil attributes other than total PTE concentration also interfere in the reproduction of collembolans. Our results support the idea that the F. candida species might not be representative of other collembolan species and that laboratory tests to assess soil contaminations should be conducted using diverse collembolan species.