The current legal regulations in species and nature conservation are mostly founded on models of ecosystems which date back to the seventies and eighties (key words: ecological equilibrium self-regulation, stability). In science, the understanding of ecosystems has profoundly changed during the last twenty years (key words: natural dynamics, stochastic change, mosaic cycle). This knowledge, however, has not yet led to a revision of conservation law. The example of the "rainforest - secondary semiarid grassland - paradoxon" documents the current existence of antagonistic conservation objectives: A similar important status is awarded to a naturally evolved high biodiversity, including species such as Titanus giganteus L. (Coleoptera: Cerambycidae) in its proper neotropical rainforest habitat on the one hand, and to an artificial species composition of comparable high diversity on man-made semiarid grassland in central Europe, including species such as Mantis religiosa (L.) (Mantodea: Mantidae) on the other hand. High species diversity and conservation in the original meaning of the word are thus the resulting aim and implementation in order to achieve sustainability of the natural resources. The consideration of the more recent findings on the functioning of ecosystems requires a different understanding of sustainability and a clear distinction among different reasons for conservation. The following three objectives should be differentiated in conservation law and other legal regulations: nature conservation to contribute to the sustainability of the natural resources, species conservation to contribute to the management of survival of individual species, and conservation of a cultural landscape to meet the socio-cultural needs. The insect diversity of secondary calcareous grasslands may be conserved for socio-cultural reasons by establishing an open-air museum, or in order to contribute to the survival of species at such a secondary site by setting up a (zoological) garden, or for nature conservation aiming at the sustainability of the natural resources by rearranging unbound dynamics including its stochastic effects. The latter, of course, would result in the local extinction of certain species and a total rearrangement of the species composition, approaching on the long term a thermophilic forest community.