Hysteresis indicates critical transition from diverse to depauperate grasslands

Background/Question/Methods

Complex systems (e.g., ecosystems, climate, financial markets) can exhibit critical transitions that are rapid, difficult to anticipate, and difficult to reverse. Critical transitions have two unique features: thresholds, where the system abruptly shifts from one stable state (or regime) to another; and hysteresis, in which the system resists shifting back to the original state even as previous conditions are restored. Although there is considerable evidence for thresholds, there is little experimental evidence for hysteresis, especially in terrestrial ecosystems. Our long-term nitrogen (N) addition and cessation experiment has been underway for three decades in a successional grassland at Cedar Creek Ecosystem Science Reserve. This design allowed us to rigorously test for two defining features of critical transitions: (1) nonlinear responses across an environmental gradient (levels of N addition), and (2) hysteresis in response to the forward and backward changes in conditions (continuous addition vs. cessation of N). Additionally, the duration of the experiment, three decades, allowed us to consider the extent to which changes were reversible.

Results/Conclusions

We found that nutrient enrichment resulted in a critical transition from a diverse to a depauperate grassland plant community by promoting a self-perpetuating invasion by an exotic species, Elymus repens. Three decades after the start of the experiment, local plant species richness was either relatively high or considerably lower. Plots that tended to have the fewest species were often dominated by E. repens. This species was either very rare or nearly completely dominant, and was rarely observed at intermediate dominance. Furthermore, E. repens dominance exhibited hysteresis in response to an experimental increase and subsequent decrease in nutrients. This implies that it could be difficult to restore biodiversity to grasslands that have undergone this critical transition. Further study will be needed to determine whether critical transitions commonly lead to declines in biodiversity and ecosystem services in grasslands, and the conditions under which these changes are predictable or reversible.