Laboratory evolution of population stability in Drosophila: constancy and persistence do not necessarily coevolve.

1. Despite considerable theoretical work, the evolution of population stability has rarely been investigated empirically. Moreover, it is not clear whether different stability properties of a population evolve together, or independently. 2. We investigate the evolution of two aspects of population stability using laboratory populations of Drosophila melanogaster selected for faster preadult development and early reproduction, and their matched controls. 3. We show that the constancy stability of the selected populations is significantly higher than their controls, confirming a previous observation that population stability can evolve as a by-product of life-history evolution. This enhanced constancy stability is due to a reduced maximal per capita growth rate, brought about by a reduction in fecundity of the selected populations as a result of the trade-off between developmental rate and fecundity. 4. Persistence stability, as reflected by the probability of extinction, does not differ significantly between selected and control populations. 5. We also show how seemingly trivial experimental details, such as the protocol for restarting extinct populations, can interact with life-history traits to alter the manifestation of the stability properties of a population.