Potential for sudden shifts in transient systems

Abstract

Thorough understanding of the potential for threshold dynamics and catastrophic shifts to occur in natural systems is of great importance for ecosystem conservation and restoration. However, verifying the presence of alternative stable states, one of the theoretical explanations for sudden shifts in natural systems, has proven to be a major challenge. We examine processes on local and landscape scales in salt-marsh pioneer zones, to assess the presence of alternative stable states in this system. To that end, we investigated the presence of typical characteristics of alternative stable states: bimodality and threshold dynamics. We also studied whether vegetation patches remained stable over long time periods. Analysis of false-color aerial photographs revealed clear bimodality in plant biomass distribution. By transplanting Spartina anglica plants of three different biomass classes on three geographically different marshes, we showed that a biomass threshold limits the establishment of Spartina patches, potentially explaining their patchy distribution. The presence of bimodality and biomass thresholds points to the presence of alternative stable states and the potential for sudden shifts, at small, within-patch scales and on short time scales. However, overlay analysis of aerial photographs from a salt marsh in The Netherlands, covering a time span of 22 years, revealed that there was little long-term stability of patches, as vegetation cover in this area is slowly increasing. Our results suggest that the concept of alternative stable states is applicable to the salt-marsh pioneer vegetation on small spatio-temporal scales. However, the concept does not apply to long-term dynamics of decades or centuries of heterogeneous salt-marsh pioneer zones, as landscape-scale processes may determine the large-scale dynamics of salt marshes. Hence, our results provide the interesting perspective that threshold dynamics may occur in systems with, on the long term, only a single stable state.