Abstract
The diversity-stability hypothesis developed over the past 25 yr appears widely misunderstood by ecologists, although it simply states that species diversity mediates community functional stability through compensating interactions to environmental fluctuations among co-occurring species. Fluctuations in the abundances of species with different adaptive modes may be a mechanism stabilizing community function in a varying environment. The available empirical evidence, including data from an experimental perturbation of successional old fields and data on the impact of environmental fluctuations on properties of grasslands in the Serengeti-Mara ecosystem of Tanzania and Kenya, suggests that the hypothesis is true at the primary producer level. As components of natural science, models are true only insofar as they are verified as accurate descriptions of the systems they purportedly characterize. The data on diversity-stability relationships in plant communities indicate that the traditional verbal model is considerably more robust in application than recent more rigorous mathematical models.