Back

Global Convergence of Plant Functional Trait Composition in the Anthropocene

Wolf, S.; Svidzinska, D.; Schellenberger Costa, D.; Mahecha, M. D.; Joswig, J.; Cai, L.; Wirth, C.; Mora, K.; Kraemer, G.; Nenoff, K.; Winter, M.; Tautenhahn, S.; Bruelheide, H.; Van Kleunen, M.; Kreft, H.; Pysek, P.; Weigelt, P.; Kattenborn, T.

2026-03-27 ecology
10.64898/2026.03.27.714725 bioRxiv
Show abstract

Since the onset of European colonial expansion, humans have accelerated species migration across continents, reshaping plant functional composition and associated ecosystem processes. Plant functional traits-such as leaf area, plant height, or rooting depth-are structured along major axes of variation, including size and leaf economics, that reflect ecological strategies. While human-mediated changes in this trait space have been documented regionally or for specific taxa, there exists no global, grid cell-level quantification of past shifts across major axes of trait variation. Here, we link global citizen science plant occurrence data with data on 37 above- and below-ground traits, and information on native and introduced status for each occurrence. Using dimension-reduction on grid cell-level trait means and introduced species status as a proxy for anthropogenic change, we identify three major axes of functional variation: the size, leaf economics, and life-span axes. By comparing past (native-only) and present-day trait distributions in 3D trait space and geographically, we find prominent region-specific shifts along all three axes. Overall, functional composition converges toward (mostly) smaller, more acquisitive, and shorter-lived assemblages, with region-specific differences in which axis shifts are most pronounced. These results provide the first global estimate of how human-mediated plant introductions have altered ecosystem functional composition in the past centuries, highlighting the spatial patterns and trait dimensions most affected by anthropogenic pressures.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.