Geographical divergence of species richness and local homogenization of plant assemblages due to climate change in grasslands

In arid and semi-arid regions worldwide, grassland plant species richness is highly sensitive to climate change. Studies assessing local grassland richness patterns have yielded inconsistent trends toward climate change, partly due to differences in recording approaches, environmental conditions, and local flora. Remote sensing presents a valuable opportunity to investigate plant richness-climate change relationships in grasslands across large environmental gradients. Based on spectral diversity indices extracted from Landsat satellite imagery, we explore how plant diversity responds to climate change and aim to determine the major climatic drivers of plant diversity patterns in ten grassland nature reserves worldwide. Plot-level plant richness was correlated with 19 bioclimatic variables through stepwise linear regression for each climate change scenario in every nature reserve. The performance of the models was assessed according to the model accuracy. We used the fitted models between climatic variables and plant richness from 1990 to 2000 to predict plant richness in 2050 and 2070 under 33 climatic change scenarios for 1120 plots in each reserve. A general tendency toward a decrease in the plot-level plant richness and beta (beta)-diversity in the future decades were observed in most cases, although there also were some opposite trends in plant richness. The dominant bioclimatic predictors involved in predictive models varied across sites. Spectral plant richness responses diverge geographically, while beta-diversity generally declines under climate change scenarios. Over the next decades, the expected homogeneities in plant species across grasslands encountered on different continents will likely lead to the dominance of climate generalist species. Policy-makers and conservationists therefore need to urgently develop strategies to ensure plant survival, particularly that of locally endemic species under predicted climatic scenarios; human assistance may be required when adjusting their distribution ranges.