Dataset DOI: 10.5061/dryad.tqjq2bwcf

Description of the data and file structure

We conducted large-scale field surveys and trait measurements across40 invaded and native plant communities spanning a 3,000 km transect in northern China to investigate how the functional strategies of Solanum rostratum vary with ecological contexts. We recorded species composition, environmental conditions, and plant functional traits to calculate CSR strategies for both invasive and native species. These data were collected to assess how functional strategy plasticity facilitates invasion success across environmental and biotic gradients.

Files and variables

1. Best_model_result.csv

Description: Model selection results for candidate linear mixed-effects models.

Variables:

• model_id — Model identifier (character)
• df — Degrees of freedom
• logLik — Log-likelihood of the model
• AICc — Corrected Akaike Information Criterion
• delta — Difference in AICc relative to the best model
• weight — AICc weight
• R2m — Marginal R² (variance explained by fixed effects)
• R2c — Conditional R² (variance explained by fixed + random effects)
• variables — Predictors included in the model (character)
• Moran_I — Moran’s I statistic for spatial autocorrelation
• Expected_I — Expected Moran’s I under the null model
• SD — Standard deviation of Moran’s I
• p_value — p-value for Moran’s I test

2. boot.csv

Description: Bootstrap results of CSR strategic dissimilarity indices (SDI).

Variables:

• SDI_C — Strategic dissimilarity index for C-strategy
• SDI_S — Strategic dissimilarity index for S-strategy
• SDI_R — Strategic dissimilarity index for R-strategy

Units: Dimensionless indices (0–1)

3. cv_dataresult.csv / cvdata.csv

Description: Trait variability data.

Variables:

• species — Species name (character)
• mean_value — Mean trait value
• sd_value — Standard deviation of trait value
• cv — Coefficient of variation
• C, S, R — CSR strategy scores for individual species (0–100%)

4. Mean_CSR_strategies.csv / CSR_strategies.csv

Description: Site-level and species-level CSR strategy scores.

Variables:

• C — Competitor strategy score (%)
• S — Stress-tolerator strategy score (%)
• R — Ruderal strategy score (%)
• Group — Species group (e.g., S. rostratum vs. native community)

5. LMMdata.csv

Description: Predictor and response variables used in linear mixed-effects models.

Variables:

• C_score_SR, S_score_SR, R_score_SR — CSR scores for S. rostratum
• C_score_CWM, S_score_CWM, R_score_CWM — Community-weighted mean CSR scores of native communities
• C_score_Dominant, S_score_Dominant, R_score_Dominant — CSR scores of dominant native species
• Grazing — Grazing intensity (categorical: grazed/ungrazed)
• Aridity — Aridity index (dimensionless)
• MAT — Mean annual temperature (°C)
• Species_richness — Number of species per plot
• Simpson — Simpson diversity index
• PD — Faith’s phylogenetic diversity (dimensionless)
• Plant_density — Plant density (individuals m⁻²)
• Biomass — Aboveground biomass (g m⁻²)
• Vegetation_coverage — Vegetation cover (%)
• MPD_phy, MNTD_phy — Mean pairwise / nearest taxon distance (phylogenetic)
• NRI, NTI — Net Relatedness Index and Nearest Taxon Index (standardized effect sizes)
• MPD_trait, MNTD_trait — Mean pairwise / nearest trait distance
• FD — Functional diversity index
• SDI_C_score, SDI_S_score, SDI_R_score — Strategic dissimilarity indices for C, S, and R
• III_SR — Invasion intensity index for S. rostratum
• Lon, Lat — Longitude and latitude of sampling site (decimal degrees)

6. random_factors.csv

Description: Site-level random effects used in LMMs.

Variables:

• Site — Site identifier (character)
• random_effect — Random effect estimate
• Lon, Lat — Site coordinates (decimal degrees)

7. model_avg_full_estimates.csv

Description: Model-averaged parameter estimates and importance values.

Variables:

• Variable — Predictor variable name
• Estimate — Model-averaged estimate
• SE — Standard error
• Adjusted_SE — Adjusted standard error
• z value — z statistic
• p — p-value
• LowerCI, UpperCI — 95% confidence interval bounds
• Significance — Significance level (e.g., ‘*’, ‘**’)
• Importance — Variable importance weight
• Frequency — Frequency of occurrence in candidate models
• Group — Variable group/category

8. StrateFy.xlsx

Description: StrateFy.xlsx is an external analytical tool originally published by Pierce et al. (2017, Functional ecology) for calculating CSR (Competitor–Stress tolerator–Ruderal) strategy scores from plant leaf traits.

This file is included in the repository solely to ensure full reproducibility of the CSR calculations performed in our study.

Important:

This Excel file is not raw data from our research.

All raw trait data used in our analyses are provided separately in unformatted CSV files.

Variables: Species ID, trait values (e.g., SLA, LDMC, LA), and site information.

Units: Trait-specific (e.g., LA in mm², LDMC in %, SLA in mm2 mg-1)).

Reference：

Pierce, S., Negreiros, D., Cerabolini, B. E. L., Kattge, J., Díaz, S., Kleyer, M., Shipley, B., Wright, S. J., Soudzilovskaia, N. A., Onipchenko, V. G., van Bodegom, P. M., Frenette-Dussault, C., Weiher, E., Pinho, B. X., Cornelissen, J. H. C., Grime, J. P., Thompson, K., Hunt, R., Wilson, P. J., … Tampucci, D. (2017). A global method for calculating plant CSR ecological strategies applied across biomes world-wide. Functional Ecology, 31(2), 444–457. https://doi.org/10.1111/1365-2435.12722

9. code_CSR.R

Description: R script containing code for trait data processing, CSR strategy calculation, SDI computation, and linear mixed-effects model fitting.

9. trait_data.csv

Description: raw leaf trait input file for matted for StrateFy

Code/software

All data processing and analyses were conducted in the free and open-source statistical environment R (version 4.3.2). The workflow is fully contained in the provided R script code_CSR.R. This script allows users to reproduce the entire analytical pipeline from raw trait data to final statistical models and figures.

Required software and packages

The following R packages were used and must be installed prior to running the script:

• StrateFy – calculation of CSR strategy scores from leaf traits.
• dplyr, tidyr – data cleaning and processing.
• ggplot2, ggtern – data visualization (e.g., ternary plots, effect plots).
• lme4 – linear mixed-effects modeling.
• MuMIn – model selection and model averaging (AICc-based).
• DHARMa – model diagnostics.
• spdep – spatial autocorrelation testing (Moran’s I).

All required packages are open source and available through CRAN.

Analytical workflow

1. CSR calculation – Leaf trait data from StrateFy.xlsx are processed to calculate C, S, and R scores for each species.
2. Trait plasticity – Within-species variation is quantified by calculating the coefficient of variation (CV) of CSR scores (cvdata.csv).
3. CSR dissimilarity – Functional divergence between invader and native communities is computed as CSR dissimilarity indices (SDI) (boot.csv).
4. Modeling and selection – Linear mixed-effects models are fitted using lme4 to assess environmental and community predictors of invasion intensity. AICc-based model selection and averaging are performed using MuMIn (Best_model_result.csv, model_avg_full_estimates.csv).
5. Model diagnostics and spatial analysis – Residual diagnostics are performed with DHARMa, and spatial autocorrelation of model residuals is tested using Moran’s I (spdep).
6. Visualization – Ternary plots of CSR strategies, CV distributions, and model effect plots are generated with ggplot2 and ggtern.

Variable Description Table

Usage

Users can open and run the script code_CSR.R directly in R. The script loads the data files automatically, executes the above steps sequentially, and reproduces the main analyses and figures reported in the manuscript.