Climate and Land Cover Change Transform Functional Identity and Reduce Functional Diversity in European Grasslands

作者：Chang Liu, Stef Haesen, Jürgen Dengler, Denys Vynokurov, Idoia Biurrun, Łukasz Kozub, Iwona Dembicz, Alireza Naqinezhad, Béla Tóthmérész, Dariia Borovyk, Franz Essl, Juan Antonio Campos, Nadiia Skobel, Péter Zsolt Török, Robin J. Pakeman, Swantje Löbel, Koenraad Van Meerbeek · 发表于：Global Ecology and Biogeography · 年份：2026 · DOI：10.1111/geb.70217 · 被引用次数：1 · 研究领域：Ecology and Vegetation Dynamics Studies、Remote Sensing in Agriculture、Land Use and Ecosystem Services

ABSTRACT Aim European grasslands rank among the most species‐rich ecosystems at small spatial scales, yet their biodiversity and functioning face significant threats from climate change and land‐use intensification. Functional traits more effectively explain ecosystem functions (EFs) than species identity or diversity. This study examines how future climate and land cover changes will shape grassland functional composition, addressing gaps in trait‐environment relationships and large‐scale functional predictions. Location Europe. Time Period 1971–2000 and 2081–2100. Major Taxa Studied 4406 distinct grassland plant species. Methods We used Boosted Regression Trees to model trait‐environment relationships based on vegetation plot data from sPlotOpen, GrassPlot, and the Nordic‐Baltic Grassland Vegetation Database (NBGVD). We mapped the 17 trait community‐weighted means (CWMs) and three functional richness (FRic) metrics under historical conditions and two future climate scenarios to assess temporal and spatial changes in grassland functional composition. Results The trait–environment relationships are highly trait‐dependent: structural and size‐related traits such as plant height, leaf area and seed number were consistently well‐predicted, whereas other traits were less well predicted. Mean annual temperature emerged as the strongest predictor of grassland functional composition. Climate and land cover change were predicted to drive significant spatial shifts in trait CWMs and F...