Data and code from: Small size species patch enhances short-term resistance to weed invasion in artificial grasslands
Data files
Jul 28, 2026 version files 52.30 KB
-
DataTwoYears_2024-2025.csv
7.75 KB
-
Figure_4_Complete.R
12.57 KB
-
Figure2_Complete.R
14.72 KB
-
Figure3_Complete.R
13.24 KB
-
README.md
4.03 KB
Abstract
Preventing weed invasion in artificial grasslands is essential for maintaining ecosystem function and sustainable management. Previous studies have shown that species patch size controls inter- and intraspecific interactions, leading to the hypothesis that it can influence weed invasion. However, there is no consensus on its actual role in regulating this process.
We manipulated four native-species patch sizes (0.0625, 0.25, 1.0 m², and mixed) to create a gradient of inter- and intraspecific interactions and tested their effect on resistance to two introduced weeds (annual Consolida ajacis and perennial Paspalum distichum). Resistance was quantified by measuring native versus invasive biomass in the artificial grasslands in two consecutive years, and key soil nutrient parameters were assessed to explore the underlying mechanisms, including soil organic carbon (SOC), ammonium (NH₄⁺), nitrate (NO₃⁻), and available phosphorus (AP).
Community biotic resistance was patch-size dependent: the smallest patches exhibited the highest resistance to invasion. Invasive biomass was negatively related to native biomass across invaded plots, indicating greater biotic resistance in more productive communities. The relationship between native productivity and biotic resistance depended on invader identity, being strong for the annual invader C. ajacis but weak for the perennial invader P. distichum.
Smaller patches were associated with higher native productivity, greater invasion resistance, and altered soil nutrient dynamics, patterns that are consistent with stronger interspecific interactions, niche complementarity and more efficient resource use for invaders. The contrasting responses of the annual and perennial weeds suggest that spatial structure interacts with invader life-history traits, with annual invaders appearing more sensitive to fine-scale competition than perennial invaders, which might be able to tolerate or exploit spatial heterogeneity at least in short-term.
This study demonstrates that spatial arrangement of native species is a key driver of invasion resistance through its effects on plant interactions and soil processes. By showing how smaller patches enhance native productivity, increase resistance to invasion and influence soil nutrient dynamics, we identify spatial biodiversity design as a practical, nature-based solution for maintaining soil health, productivity, and short-term resilience in artificial grasslands. These findings advance ecological understanding by linking spatial community structure with above–belowground feedbacks that govern ecosystem stability.
Dataset DOI: 10.5061/dryad.mw6m906c6
Description of the data and file structure
Small size species patch enhances short-term resistance to weed invasion in artificial grasslands
This dataset supports the research titled Small size species patch enhances short-term resistance to weed invasion in artificial grasslands, which focuses on how native species patch size regulates biotic resistance to weed invasion in artificial grasslands over two consecutive years (2024–2025).
We collected plot-level aboveground biomass of native and invasive plants, as well as soil physicochemical properties including soil organic carbon, nitrate nitrogen, ammonium nitrogen, available phosphorus and soil pH.
Core experimental results: Smallest native species patches exhibited the highest biotic resistance to weed invasion, and invasive biomass was negatively correlated with native plant biomass across invaded plots. The strength of such relationship depended on invader life-history strategies: significant for annual invader C. ajacis and weaker for perennial invader P. distichum. Patch size modified native community productivity and soil nutrient status, jointly shaping invasion outcomes.
Files and variables
File: DataTwoYears_2024-2025.csv
Description: Raw and processed experimental dataset for 2024–2025 field sampling, including plant biomass and soil nutrient data
Variables
- BlockID: Randomized block identifier
- PlotID: Unique experimental plot ID
- Year: Sampling year
- Rep: Treatment replicate number
- Patch_size: Native species spatial arrangement treatment
- Invader: Invasion treatment
- NativeAB: Native plant aboveground biomass
- InvasiveAB: Invasive weed aboveground biomass
- InvPropAB: Proportion of invasive biomass within total plot biomass
- PlotAB_gm2: Total plot aboveground biomass (native + invasive)
- pH: Soil pH value
- NO3: Soil nitrate nitrogen
- NH4: Soil ammonium nitrogen
- AP: Soil available phosphorus
- SOC: Soil organic carbon
File: Figure_4_Complete.R
Description: Soil nutrient analysis and PCA ordination
File: Figure2_Complete.R
Description: Total biomass comparison and native–invasive biomass relationships
File: Figure3_Complete.R
Description: Native productivity and biotic resistance analysis
Abbreviation definition
SM: Small patch
MP: Medium patch
LP: Large patch
Mixed: Uniform mixed sowing treatment
Code/software
All statistical analyses were performed in R (version 4.2.0, R Core Team). Because the same experimental plots were measured in both 2024 and 2025, responses were analysed using linear mixed-effects models fitted with the lme4 and lmerTest packages.
Given the unbalanced replication structure and repeated-measures design, significance of fixed effects was assessed using Type III analysis of variance with Satterthwaite approximations for denominator degrees of freedom. Model assumptions of normality and homoscedasticity were assessed through visual inspection of residuals and diagnostic plots. When significant main or interaction effects were detected, pairwise comparisons were performed using the emmeans package, with Tukey-adjusted multiple comparisons. Results are presented as estimated marginal means ± standard errors unless otherwise stated.|
Access information
Other publicly accessible locations of the data:
- The complete dataset is hosted on the Dryad Digital Repository: https://doi.org/10.5061/dryad.mw6m906c6
Data was derived from the following sources:
- All data presented in this collection were directly measured and collected by the authors during a field experiment at Baima Farm experimental site, Nanjing, China.
