Abrupt transition of mesic grassland to shrubland: evidence for thresholds, alternative attractors, and regime shifts. Issue 9 (1st September 2014)

• Record Type: Journal Article
• Authors: Ratajczak, Zak; Nippert, Jesse B.; Ocheltree, Troy W.
• Is Part Of: Ecology. Volume 95:Issue 9(2014)
• Page Start: 2633
• Page End: 2645
• Publication Date: 2014-09-01
• Subjects: alternative stable states -- bi-stability -- bush encroachment -- catastrophic shifts -- critical transitions -- dynamical systems -- fold bifurcation -- grassland–woodland transitions -- shrub encroachment -- tallgrass prairie -- tipping points -- woody encroachment; Ecology -- Periodicals; Ecology -- Periodicals; Écologie -- Périodiques; Ecologie; Écologie; Écologie animale; Écologie végétale; Ecology; Periodicals; 577.05
• DOI: 10.1890/13-1369.1
• Languages: English
• ISSNs: 0012-9658

Abstract

Abstract : Ecosystems with alternative attractors are susceptible to abrupt regime shifts that are often difficult to predict and reverse. In this study, we quantify multiple system dynamics to determine whether the transition of mesic grassland to shrubland, a widespread phenomenon, represents a linear reversible process, a nonlinear but reversible threshold process, or a transition between alternative attractors that is nonlinear and prone to hysteresis. Using a 28‐yr data set with annual resolution and extensive spatial replication, we found that shrub cover is correlated with distinct thresholds of fire and C4 grass cover, resulting in temporal bimodality of shrub cover and abrupt shifts of shrub cover despite gradual changes in grass cover. These abrupt increases in shrub cover are the most rapid ever reported in grasslands, and illustrate internal thresholds that separate grasslands and shrublands. Nonlinear transitions from low to high shrub cover were also closely associated with positive feedback mechanisms that alter fire and competition ( r 2 = 0.65), suggesting that grasslands and shrublands could show hysteresis, and by definition exist as alternative attractors. Thus, the response of this ecosystem to anthropogenic activity should tend to be rapid, nonlinear, and perhaps difficult to reverse. Regime shifts in this mesic grassland were predictable: we found that grassland and shrubland attractors were differentiated by critical thresholds of ∼50–70% grass cover, 5–10% shrub cover, and a fire return interval of ∼3 yr. These thresholds may provide adaptive potential for managing nonlinear behavior in socio‐ecological systems in a changing environment.