Skip to main content

Selection on convergent functional traits drives compositional divergence in a tallgrass prairie restoration experiment

Cite this dataset

Karimi, Nisa; Hipp, Andrew (2021). Selection on convergent functional traits drives compositional divergence in a tallgrass prairie restoration experiment [Dataset]. Dryad.


1. Plant biodiversity is often partitioned into taxonomic diversity (species composition and abundance), phylogenetic diversity (breadth of evolutionary lineages) and functional diversity (resource‐use strategies or physical traits). Evaluating the effects and interplay of these dimensions can provide insights into how assembly processes drive compositional changes in plant communities. However, teasing apart the effects of different biodiversity dimensions is challenging in observational studies or retrospective analyses.

2. To evaluate how plant phylogenetic and trait history shape community establishment and turnover in restoration of a species‐rich North American tallgrass prairie, we conducted an experiment with 127 species planted in assemblages representing three levels of phylogenetic diversity (PD) and two of functional trait diversity (FD), holding starting species richness (SR) fixed. We tested whether PD and FD of planted assemblages predicted species diversity, compositional turnover and selection on functional traits.

3. Rank order of initial functional and phylogenetic diversity levels was maintained throughout the experiment, but neither diversity measure correlated positively with species richness by the end of the experiment. Phylogenetic and taxonomic beta diversity increased among all treatments. This increase in compositional beta diversity was associated with directional selection on phylogenetically dispersed functional traits. A set of functional traits associated with competitiveness in tallgrass prairies predicted species' cover for all survey years: stem dry matter content, leaf dry matter content, vegetative height and rhizomatous growth. Although all plots collectively converged on a similar suite of functional traits, functional beta diversity increased among high‐FD plots.

4. Synthesis. Neither higher functional nor phylogenetic diversity maintained higher species richness (SR) over time in our study. Although SR was not maintained, higher levels of PD and FD were. Both types of diversity shaped the rate at which plots changed in composition over time, with high diversity treatment plots increasing in beta diversity. Selection for traits convergent across the tree of life drove phylogenetic and compositional divergence among plots. While optimization of site‐specific functional traits may be most important for maintaining higher SR, our work implies that planting higher initial PD and FD may make grassland restorations more adaptable to site conditions that may be difficult to predict.


Data was collected from a mesic tallgrass prairie restoration experiment located at The Morton Arboretum in Lisle, Illinois, USA (41.821°N, 88.093°W). The experiment is a fully factorial phylogenetic diversity (PD) × functional diversity (FD) experiment with three PD levels (low, medium and high) and two FD levels (low and high) consisting of a total of 127 species characteristic of mesic eastern tallgrass prairies. Vegetative cover was collected twice annually. Plant functional traits were gathered for this experiment and from elsewhere (see manuscript).

Usage notes

R scripts are labeled for order of use. Not all coulmns of the spreadsheets provided are used in this study; please refer to the readme.txt, manuscript for details and R code directly.


National Science Foundation, Award: DEB 1354551

National Science Foundation, Award: DEB 1354426