Inter-individual variation promotes ecological success of populations and species: evidence from experimental and comparative studies

Biological diversity is threatened by exploitation, fragmentation of natural habitats, pollution, climate change, and anthropogenic spread of species. The question of how among-individual variation influences the performance of populations and species is a poorly explored but currently growing field of research. Here, we review 31 experimental and 14 comparative studies and first investigate whether there is empirical support for the propositions that higher levels of among-individual phenotypic and genetic variation promote the ecological and evolutionary success of populations and species in the face of environmental change. Next, we examine whether and how the effect of diversity depends on environmental conditions. Finally, we explore whether the relationship linking population fitness to diversity is typically linear, asymptotic, or whether the benefits peak at intermediate diversity. The reviewed studies provide strong, almost invariable, evidence that more variable populations are less vulnerable to environmental changes, show decreased fluctuations in population size, have superior establishment success, larger distribution ranges, and are less extinction prone, compared with less variable populations or species. Given the overwhelming evidence that variation promotes population performance, it is important to identify conditions when increased variation does not have the theoretically expected effect, a question of considerable importance in biodiversity management, where there are many other practical constraints. We find that experimental outcomes generally support the notion that genetic and phenotypic variation is of greater importance under more stressful than under benign conditions. Finally, population performance increased linearly with increasing diversity in the majority (10 of 12) of manipulation studies that included four or more diversity levels; only two experiments detected curvilinear relationships.