Understanding the relationships between vegetation dynamics (greening and browning) and resilience (the capacity to recover from disturbances) is key to effective vegetation restoration and management. However, empirical evidence for their relationships has not been well investigated and critical questions remain about whether and how vegetation greening may contribute to vegetation resilience to disturbances. Here, using a recently proposed vegetation index (kNDVI) and a resilience metric (the autoregressive coefficient of a first-order model, AR(1)), we assessed vegetation dynamics and resilience changes from 2000 to 2020 in China’s Loess Plateau (LP). We found that although the LP showed an overall trend of vegetation greening, vegetation resilience did not increase concurrently. Our analysis also revealed that a larger area of vegetation showed a signal of resilience loss in the second decade 2011-2020 (42.45% of LP’s vegetated land) than in the first decade 2000-2010 (20.22%). Vegetation density (i.e., changes in kNDVI) was the most important explanatory variable of resilience change, and greening (i.e., increase in kNDVI) overall contributed to resilience gains. However, this positive effect of vegetation greening on resilience was partly offset by climate factors including increasing mean temperature and precipitation variability in the second decade (2011-2020), which explained the widespread resilience loss in this period. Combined with the results and an adaptive cycle model, we showed that the evolution of vegetation ecosystems of the LP has shifted from the growth stage (r) to the conservation stage (K), characterized by increasingly locked resources (e.g., biomass, nutrients, and water) in vegetation and therefore less flexible and resilient to external shocks. To ensure the sustainability of vegetation ecosystems, we highlight the need to integrate resilience indicators into ecological restoration and conservation efforts.
 

植被动态，植被弹性，早期预警信号，适应性循环模型，黄土高原