Synthesis of scale-dependent plant biodiversity change in climate warming experiments

Climate warming is one of the most pressing environmental challenges for plant populations, communities, and ecosystems on earth. Warming experiments are an important tool for establishing causal relationships between climate change and plant diversity responses. However, results from such experiments vary in direction and magnitude. We hypothesize that such variation arises from context-dependent factors, such as the magnitude of the temperature manipulation, the study duration, the spatial grain of the analysis, the background climate in which the study was conducted, and the biodiversity metrics used. Here, we present a comprehensive synthesis of primary data from 42 warming experiments, mostly from tundra and grassland ecosystems. We examined how warming influenced composition and how responses of plant diversity varied at the alpha (= plot), beta (= plot-to-plot), and gamma (= site) scales. Context-dependent factors explained some of the variation observed in responses across studies. Higher magnitudes of warming led to stronger declines in species richness at the alpha scale. Experimental warming led to stronger declines in species richness in plant communities with warmer background summer temperature (e.g., desert grasslands) compared to those in communities exposed to colder background summer temperatures (e.g., tundra communities). Warming decreased evenness at the local (alpha) scale, but increased evenness at the beta scale, indicating that replicate warming plots diverged in their abundance distribution over time. Warming changed the composition of plant communities, particularly through species turnover rather than in the nestedness component, suggesting that warming leads to novel species assemblages. Our study highlights that responses of terrestrial plant communities to warming depend on experimental methodology, environmental context, and the spatial grain of investigation.