A major challenge to balancing natural resource use with ecosystem conservation is the difficulty of restoring or rehabilitating heavily exploited, severely damaged environments. Biotic communities introduced to extensively altered landscapes can face high stress and stochasitc perturbations that destabilize community properties and revert "restored" ecosystems back to degraded states. One potential control of resistance to such backsliding is the level of biodiversity introduced to the environment by managers. While much-studied in ecology, relationships between biodiversity and stability properties have rarely been tested in the context of restoration or high-stress ecosystems. To understand the relationship between introduced biodiversity and the capacity for restored communities to withstand extreme environmental conditions, we conducted a biodiversity addition experiment on high-stress abandoned limestone quarry floors. We grew 8 forb and grass species characteristic of rare natural limestone pavements called alvars in greenhouse monocultures and then assembled 256 pot communities of equal plant density and cover, using 1, 2, 4 or 8 species. Communities were composed using a “broken stick” design that established all species with equal abundance at each richness level, enabling separation of species identity and interaction effects. In early spring, after 6 months of greenhouse growth, we transplanted the communities to plots on 4 limestone quarry floors and monitored community and population-level changes in plant cover over a 132-day period.

Results/Conclusions

Heat, drought, and poor soil resources on quarry floors led to steeply reduced plant performance over time. However, at the community level reduced vegetation cover was almost exclusively attributable to communities with 1-4 species, as cover in 8-species communities actually increased slightly. At the population level, a similar positive influence of biodiversity was detected: 5 species were strongly stabilized by increased neighbour richness, 1 species was weakly stabilized, and 2 species were negligibly influenced. Increased resistance with richness at both population and community levels is consistent with facilitation reducing stress and perturbation impacts, but not with compensatory species dynamics underlying stability. The chief implication of our results is that by introducing a high diversity of positively-interaction species, managers of high-stress ecosystems can maximize the stability of restored communities and enhance the success of difficult restoration projects.