Dominant species predict plant richness and biomass in global grasslands

Zhang, Pengfei and Seabloom, Eric W. and Foo, Jasmine and MacDougall, Andrew S. and Harpole, W. Stanley and Adler, Peter B. and Hautier, Yann and Eisenhauer, Nico and Spohn, Marie and Bakker, Jonathan D. and Lekberg, Ylva and Young, Alyssa L. and Carbutt, Clinton and Risch, Anita C. and Peri, Pablo L. and Smith, Nicholas G. and Stevens, Carly J. and Prober, Suzanne M. and Knops, Johannes M. H. and Wardle, Glenda M. and Dickman, Christopher R. and Ebeling, Anne and Roscher, Christiane and Martinson, Holly M. and Martina, Jason P. and Power, Sally A. and Niu, Yujie and Ren, Zhengwei and Du, Guozhen and Virtanen, Risto and Tognetti, Pedro and Tedder, Michelle J. and Jentsch, Anke and Catford, Jane A. and Borer, Elizabeth T. (2025) Dominant species predict plant richness and biomass in global grasslands. Nature Ecology and Evolution. ISSN 2397-334X

Text (Homogenization_25_indicating changes)
Homogenization_25_indicating_changes.pdf - Accepted Version
Available under License Creative Commons Attribution.
Download (287kB)

Abstract

The bidirectional relationship between plant species richness and community biomass is often variable and poorly resolved in natural grassland ecosystems, impeding progress in predicting impacts of environmental changes. Most biological communities have long-tailed species abundance distributions (for example, biomass, cover, number of individuals), a general property that may provide predictive power for species richness and community biomass. Here we show mathematical relationships between community characteristics and the abundance of dominant species arising from long-tailed distributions and test these predictions using observational and experimental data from 76 grassland sites across 6 continents. We find that community biomass provides little predictive ability for community richness, consistent with previous findings. By contrast, the relative abundance of dominant species quantitatively predicts species richness, whereas their absolute abundance quantitatively predicts community biomass under both ambient and altered environmental conditions, as expected mathematically. These results are robust to the type of abundance measure used. Three types of simulated data further show the generality of these results. Our integrative framework, arising from a few dominant species and mathematical properties of species abundance distributions, fills a persistent gap in our ability to predict community richness and biomass under ambient and anthropogenically altered conditions.