Stress gradients interact with disturbance to reveal alternative states in salt marsh: Multivariate resilience at the landscape scale

Abstract

Abstract Stress gradients influence many ecosystem processes and properties, including ecosystem recovery from and resistance to disturbance. While recent analytical approaches have advanced multivariate metrics of ecosystem resilience that allow quantification of conceptual resilience models and identification of thresholds of state change, these approaches are not often translated to landscape scales. Using natural and restored salt marshes in Louisiana, USA, we quantified plant community recovery and resistance metrics along flooding stress gradients. n‐dimensional hypervolumes of plant community biomass and structure were simulated using field data collected from disturbance‐recovery experiments. The relationships between multivariate resilience metrics and flooding stress gradients were then mapped at community‐ and landscape‐relevant scales by scaling with airborne‐derived data across the region. Greater pre‐disturbance abiotic stress decreased live below‐ground, but not above‐ground, biomass, and ultimately led to lower post‐disturbance total recovery, recovery rates and resistance of plant communities. Vegetated plots flooded >52% of the time transitioned to an alternative, unvegetated state after disturbance. Mapping revealed differences in spatial patterns of resilience—highlighting low, interior marsh edges as especially vulnerable to the combination of chronic flooding stress and acute disturbance. At the landscape scale, approximately half of the area (48%) is vulnerable to state change after pulse disturbances. Synthesis. Ultimately, we quantify the ball‐and‐cup conceptual model for a salt marsh ecosystem and its alternative state, mudflat. We find that increasing abiotic stress due to climate change diminishes ecosystem resilience, but the interaction with common episodic disturbances is necessary to reveal transitions to alternative states and quantify state change thresholds. Quantifying and mapping resilience and where alternative states may exist in this fashion improves ecologists' ability to investigate the mechanisms of stress gradient control on emergent ecosystem properties while providing spatially explicit resources for managing ecosystems according to their projected resilience.