Effects of biodiversity on ecosystem stability: distinguishing between number and composition of species

Abstract

Declines in biodiversity have caused concern because of ethical and aesthetic reasons, but also because of the consequences for the goods and services provided by natural ecosystems. Consequently, ecologists have focused for decades on testing the idea that systems with more species are more stable. The results, however, have been complex and inconsistent. In particular, it is still unclear whether high stability in species-rich communities is due to the number of species per se (species richness) or to the increased likelihood of including particular species or functional types (species composition). In this thesis, I evaluated the contribution of species richness and species identity to the stability of marine hard-bottom communities. Combining observational and manipulative experimental methods, I conducted three field studies in intertidal and shallow subtidal habitats of Helgoland Island, NE Atlantic. First, I conducted an observational study to test whether intertidal communities containing many species are more stable (i.e. do vary less over time) than communities containing fewer species. Species covers were estimated every 6 months for 24 months and an index of stability was calculated for total community cover across time (S = mean SD-1). Second, I conducted a synthetic-assemblage experiment - in which I increased the diversity of field-grown sessile suspension-feeding invertebrates - to determinate whether assemblages containing several functional groups consume a greater fraction of resources than is caught by any of the functional types grown alone. (A functional group is a group of species with the same effect on an ecosystem property.) Finally, I conducted a removal experiment to test whether the loss of the canopy-forming alga Fucus serratus and mechanical disturbances that provide free substratum affect the temporal variability in cover of intertidal communities. In the removal experiment, species covers were estimated every 3 months for 18 months and the temporal variance was analysed.In general, the effects of the number of species and functional groups on ecosystem stability were weaker than those of species composition. In the observational study, stability was a negative and curvilinear function of species richness, which probably resulted from the dominance of few species. In accordance, the synthetic-assemblage experiment showed that there was no relationship between resource consumption and functional group diversity per se, but that different functional groups had idiosyncratic effects. On the other hand, the removal of Fucus changed the physical environment by increasing temperature, irradiance, and amount of sediment, which depressed the abundance of sensitive species like encrusting algae and small sessile invertebrates, but raised the abundance of more tolerant species like ephemeral green algae. This resulted in a significant increase in the variability of species abundances, but not in that of communities. The negative covariances resulting from the compensation between sensitive and tolerant species buffered the community stability against the environmental disturbances. These patterns were consistent across two sites, suggesting a consistent effect of canopies across the spatial variability of this system.Species composition appears to be more important for ecosystem stability than taxonomic and functional richness. Yet, the occurrence of compensatory dynamics in the face of environmental changes (i.e. the removal of Fucus) suggests that a variety of species with differing environmental tolerances is needed to maintain the functioning of this ecosystem. Therefore, predicting the consequences of species loss requires a detailed knowledge about the effects of species on ecosystem functioning and their responses to the environment. Conservational managers should strive (i) in identifying species with disproportional effects on ecosystem functioning, and (ii) in maintaining a redundancy of species with similar effects on ecosystem functioning and a diversity of species with different sensitivities to a suite of environmental conditions.