Background/Question/Methods Investigating the mechanisms responsible for ecological thresholds is essential to understanding processes leading to ecosystem regime shifts. Dryland ecosystems are especially prone to threshold behavior wherein stressor-mediated alteration of patterns and processes can shift systems to alternative states resulting in compromised ecosystem services (e.g., maintenance of biological diversity and soil stability). Despite wide acceptance of the concept, few studies have used or quantified ecological thresholds for direct application to land management. By understanding and quantifying indicators of threshold behavior, land managers may be able to predict an ensuing threshold crossing and implement management actions to prevent a transition to persistent degraded conditions. We are using a combination of observational and experimental data, conceptual and simulation models, and meta-analyses to explore threshold dynamics across multiple ecological sites in dryland ecosystems of the Colorado Plateau region. A primary objective of this research is to determine indicators which (1) provide early warnings of ecosystem degradation in advance of a threshold crossing and (2) can be practically monitored by land management agencies of the Colorado Plateau region.   

Results/Conclusions Key degradational processes characteristic of dryland ecosystems include disturbance of biological soil crusts (BSCs), decreased soil stability and increased erosion, and invasion by nonnative plant species. Based on data from sites experiencing various disturbance intensities, we classified 10 biotic and abiotic degraded “syndromes” indicative of potential thresholds. Several syndromes were observed along axes using multivariate ordination. We detected nonlinear relationships between cover of BSCs and soil stability, with soil stability becoming unstable below thresholds in BSC cover. We used simulation models to understand the consequences of decreasing soil stability, and found wind erosion thresholds depend on interactions between soil stability, vegetation structure, and wind speed. Similarly, results suggest that invasions by nonnative plants and compositional shifts depend on interactions between soil properties, land use, proximity to invader propagules, and climate. These results illustrate that whereas general factors leading to degraded states or syndromes can be identified across regions, knowledge of specific mechanisms of threshold behavior for individual sites requires understanding interactions among multiple biotic and abiotic variables across spatial scales. We suggest that early warning indicators of regime shifts to degraded states must include a multivariate approach to quantifying threshold values for use by land management agencies.