When species partition the temporal dimension through their responses to a changing physical environment, species coexistence may occur by the storage effect. Such coexistence relies on storage in a population of the fruits of favorable periods. It requires also covariance over time between environmental favorability and resource or apparent competition. Partitioning of time means that this covariance is predicted to be positive while a species is at high density, but is predicted to fall to lower values if density falls. Positive covariance between environment and competition inhibits population increases during favorable periods. It is the lessening of this inhibition as density falls that promotes recovery from low density, and hence coexistence by temporal partitioning. Detecting change in covariance with a change in density is thus key to definitive testing of coexistence by temporal partitioning.

Results/Conclusions

Several different sorts of experimental and observational studies might be used to detect covariance change. The ideal study is an invader-resident comparison where covariance in the invader state, in which a species is depressed to low density, is compared with covariance in the resident state, in which the species has not been manipulated. A second type of study involves species removals, and compares covariance between monoculture and polyculture. Covariance between environment and competition might be measured directly, or alternatively its change inferred by its predicted effects on fluctuations in population growth rates or recruitment. These various approaches to testing the storage effect will be illustrated with data from field studies of the herbaceous understorey of an Australian eucalypt forest, annual plants in southwestern USA, and desert ephemerals of South Australia.