Experimental warming drives local grassland plant species loss

Climate change poses a growing threat to many ecosystems, including grasslands, which are a current priority for conservation due to their vulnerability to interacting threats from human activity. North American grasslands are expected to experience warmer temperatures and more frequent and severe droughts in the coming decades, with potential consequences for native biodiversity. We conducted an experiment at Cedar Creek Ecosystem Science Reserve, Minnesota, USA, to investigate how warming and drought treatments affected grassland plant community structure over six years in plots planted with species mixtures. Warming consistently reduced plant species richness, with its effects on Shannon diversity (which additionally considers species’ relative abundances) and dominance varying across years. These warming-by-year interactions were likely driven by temporal variability in environmental conditions and species-specific responses. Notably, legumes consistently showed positive responses to warming. Drought alone had minimal direct effects on species richness or diversity but reduced variability in diversity responses over time, suggesting greater stability of diversity under drought conditions. This study underscores the important role of warming in reducing species richness, altering diversity, and reshaping functional group composition in grassland ecosystems. While temporal variability influenced the magnitude of warming effects on diversity, legumes’ positive responses highlight the importance of functional group dynamics in potentially buffering against species loss. Long-term experiments that allow consideration of interannual variability are essential for improving predictions of ecosystem responses and informing adaptive management strategies aimed at sustaining biodiversity and ecosystem functioning in grasslands.

Data structure and concepts: This dataset contains experimental and environmental data collected from a multi-year BACD (Biodiversity and Climate with Drought) experiment, which applies warming and drought manipulation in grassland plots (2017–2022) at Cedar Creek Ecosystem Science Reserve. Data are provided in several files capturing different community and ecosystem-level responses: species presence/absence, species-level biomass (including log-transformed values), community-level diversity and dominance indices, functional group biomass proportions, daily climate summaries, standardized precipitation–evapotranspiration index (SPEI) values, and subplot-level soil moisture responses. Data values are primarily continuous (e.g., biomass in g/m², soil moisture in m³/m³, daily temperature and precipitation, diversity indices) or categorical (e.g., treatment type, year, subplot). The dataset is broadly reusable for meta-analyses of global change experiments, ecological modeling and parameterization, and teaching applications. No sensitive information is included, and no legal or ethical restrictions apply to reuse beyond standard citation and attribution.