Transition from global stability to multiple attractors in microcosms

Species-rich microbial communities have recently become an experimental proving ground for a long-standing theoretical idea – that ecological dynamics can emerge predictably from the multiplicity of species interactions, rather than from specific biological traits or functions. This theoretical picture predicts characteristic patterns relating community diversity, stability and invasions – patterns that have recently been observed in microcosms. However, one key aspect of this picture that was not directly tested is the potential for each community to exhibit multistability, a dynamical property whose ecological consequences range from history dependence to catastrophic regime shifts. Here we assembled ~100 bacterial communities, manipulating their species pool and interaction strength, and tested each community’s dynamics from various initial species abundances. Our experiments confirm a central theoretical prediction: as species pool size and mean interaction strength increase there is a transition from a single globally stable equilibrium to a multiplicity of attractors, where the same community can reach multiple stable or fluctuating attractors under identical environmental conditions. Besides this complexity-driven transition, we also uncovered a biologically driven mechanism for alternative stable states: multi-species communities formed two distinct clusters, an acidic regime with low biomass and an alkaline regime with high biomass, suggesting the possibility of abrupt shifts in community-level functioning. Some communities reached both acidic and alkaline regimes while also exhibiting compositional multistability within a given regime, a phenomenon we term “hybrid multistability”. Understanding high-diversity ecosystems therefore requires a novel perspective combining complexity-driven phenomena with key large-scale biological drivers.