Repository containing field data, Excel files, and R scripts for the analyses in "Thicket Formation and Light Limitation Drive Patch-Scale Vegetation Transitions in a Semi-Arid Savanna."

This repository contains field-collected vegetation and environmental data, Excel spreadsheets, R scripts, and output figures used in the analyses for the manuscript “Thicket Formation and Light Limitation Drive Patch-Scale Vegetation Transitions in a Semi-Arid Savanna.” The dataset includes woody and herbaceous species composition, canopy cover, light transmittance, and leaf area index (LAI) data collected from 51 sites along a woody encroachment gradient at Endwell Farm, Eastern Cape Province, South Africa. Historical aerial imagery from 1949–2021 was used to assess woody cover changes, complemented by cluster analysis, indicator species analysis, and quantile regression. The repository also includes README files describing data structure and instructions for running the R scripts. Each dataset includes a dedicated README Word document explaining variable definitions and data structure: File Description Readme_CoverLightTrees.docx Describes columns and variables in CoverLightTrees.xlsx. Readme_Env.docx Describes variables in Env.xlsx. Readme_indicator_values(all).docx Explains outputs from the IndVal (Indicator Species) analysis. Readme_Matrices.docx Describes structure and variable coding of species abundance matrices (MatrixHerb, MatrixTree, and MatrixTreeHerb). R Scripts (used to generate figures in the manuscript) Each R file corresponds to a figure or analysis step described in the paper: R Script Description Figure_3_woody_cover_changes.R Analysis and visualization of woody cover changes over time (1949–2021). Figure_4_Table1_indiciespecies.R Indicator species analysis (IndVal) and cluster characterization. Figure_5_herb_sankey_and_tree_sankey.R Sankey diagrams visualizing connections between herbaceous and tree communities across encroachment stages (does not require Excel input; uses IndVal output). Figure_6_Herbaceous_basal_cover.R Quantile regression and visualization of basal cover responses to woody cover and transmittance. Excel Spreadsheets (raw and processed data) File Description CoverLightTrees.csv Field measurements of tree cover, light transmittance, and leaf area index (LAI). Env.csv Environmental variables (e.g., light transmittance, LAI, site metadata). MatrixHerb.csv Species abundance matrix for herbaceous vegetation across 51 sites. MatrixTree.csv Species abundance matrix for woody species across 51 sites. MatrixTreeHerb.csv Combined abundance matrix of tree and herbaceous species used for cluster analysis. Tree_cover_changes.csv Long-term changes in woody cover derived from aerial imagery (1949–2021).