Aridity-dependent shifts in biodiversity-stability relationships but not in underlying mechanisms.

Climate change will affect the way biodiversity influences the stability of plant communities. Although biodiversity, associated species asynchrony, and species stability could enhance community stability, the understanding of potential nonlinear shifts in the biodiversity-stability relationship across a wide range of aridity (measured as the aridity index, the precipitation/potential evapotranspiration ratio) gradients and the underlying mechanisms remain limited. Using an 8-year dataset from 687 sites in Mongolia, which included 5496 records of vegetation and productivity, we found that the temporal stability of plant communities decreased more rapidly in more arid areas than in less arid areas. The result suggests that future aridification across terrestrial ecosystems may adversely affect community stability. Additionally, we identified nonlinear shifts in the effects of species richness and species synchrony on temporal community stability along the aridity gradient. Species synchrony was a primary driver of community stability, which was consistently negatively affected by species richness while being positively affected by the synchrony between C3 and C4 species across the aridity gradient. These results highlight the crucial role of C4 species in stabilizing communities through differential responses to interannual climate variations between C3 and C4 species. Notably, species richness and the synchrony between C3 and C4 species independently regulated species synchrony, ultimately affecting community stability. We propose that maintaining plant communities with a high diversity of C3 and C4 species will be key to enhancing community stability across Mongolian grasslands. Moreover, species synchrony, species stability, species richness and the synchrony between C3 and C4 species across the aridity gradient consistently mediated the impacts of aridity on community stability. Hence, strategies aimed at promoting the maintenance of biological diversity and composition will help ecosystems adapt to climate change or mitigate its adverse effects on ecosystem stability.

Dryland sensitivity to climate change and variability using nonlinear dynamics.

Primary productivity response to climatic drivers varies temporally, indicating state-dependent interactions between climate and productivity. Previous studies primarily employed equation-based approaches to clarify this relationship, ignoring the state-dependent nature of ecological dynamics. Here, using 40 y of climate and productivity data from 48 grassland sites across Mongolia, we applied an equation-free, nonlinear time-series analysis to reveal sensitivity patterns of productivity to climate change and variability and clarify underlying mechanisms. We showed that productivity responded positively to annual precipitation in mesic regions but negatively in arid regions, with the opposite pattern observed for annual mean temperature. Furthermore, productivity responded negatively to decreasing annual aridity that integrated precipitation and temperature across Mongolia. Productivity responded negatively to interannual variability in precipitation and aridity in mesic regions but positively in arid regions. Overall, interannual temperature variability enhanced productivity. These response patterns are largely unrecognized; however, two mechanisms are inferable. First, time-delayed climate effects modify annual productivity responses to annual climate conditions. Notably, our results suggest that the sensitivity of annual productivity to increasing annual precipitation and decreasing annual aridity can even be negative when the negative time-delayed effects of annual precipitation and aridity on productivity prevail across time. Second, the proportion of plant species resistant to water and temperature stresses at a site determines the sensitivity of productivity to climate variability. Thus, we highlight the importance of nonlinear, state-dependent sensitivity of productivity to climate change and variability, accurately forecasting potential biosphere feedback to the climate system.