Abstract
Predicting biodiversity and dynamics of complex communities is a fundamental challenge in ecology. Leveraging bacterial microcosms with well-controlled laboratory conditions, Hu et al. recently performed a direct test of theory predicting that two community-level parameters (i.e., species pool size and inter-species interaction strength) dictate transitions between three dynamical phases: stable full coexistence, stable partial coexistence, and persistent fluctuations. Generally, communities experience species extinctions before they lose stability as either of the two parameters increases.