Dataset for: Coupled decline in growth and fine-branching signals the onset of drought-induced European beech mortality

Code and data for “Coupled decline in fine-branching and growth signals the onset of drought-induced European beech mortality" This archive contains all data and code required to reproduce the analyses and figures presented in the manuscript. All scripts are fully commented and designed to be run in a defined order (see below). TLS data, as well as growth and vitality data are zipped files that need to be extracted prior to usage. Metadata Explaining the rows and columns from the vitality_survey_intense.csv and growth_data_intense.csv datasets (see also below). Main scripts (R) growth.RPreparing growth data for subsequent analyses (growth curve estimation, drought-induced growth reductions). tree_qsm.RCorrections of QSM models, visualisation and derivation of tree metrics. statistical_analyses.RBayesian statistical analyses and modelling, including predictions and visualisation of results. Recommended workflow (script execution order) To reproduce the full workflow, run scripts in the following order: growth.R tree_qsm.R statistical_analyses.R The output generated by earlier scripts is used as input for later scripts. Folder structure . ├── README.md ├── metadata.csv ├── growth.R ├── tree_qsm.R ├── statistical_analyses.R ├── tls_data/ │ ├── 2214/ │ │ ├── 1_pointclouds/ │ │ ├── 2_treegraph_qsm/ │ │ └── 3_rtwig_adjusted_qsm/ │ ├── 2215/ │ │ ├── 1_pointclouds/ │ │ ├── 2_treegraph_qsm/ │ │ └── 3_rtwig_adjusted_qsm/ │ ├── 2218/ │ │ ├── 1_pointclouds/ │ │ ├── 2_treegraph_qsm/ │ │ └── 3_rtwig_adjusted_qsm/ │ ├── 2219/ │ │ ├── 1_pointclouds/ │ │ ├── 2_treegraph_qsm/ │ │ └── 3_rtwig_adjusted_qsm/ │ ├── 2220/ │ │ ├── 1_pointclouds/ │ │ ├── 2_treegraph_qsm/ │ │ └── 3_rtwig_adjusted_qsm/ │ ├── 2221/ │ │ ├── 1_pointclouds/ │ │ ├── 2_treegraph_qsm/ │ │ └── 3_rtwig_adjusted_qsm/ └── growth_and_vitality_data/ ├── growth_data_intense.csv └── vitality_survey_intense.csv Data description 1) Terrestrial laser scanning (TLS) data (tls_data/) Scanning instrument: We used a RIEGL VZ-400i with a wavelength of 1550 nm, 0.35 mrad beam divergence, and pulse repetition rate of up to 1200kHz. The folder tls_data/ contains all terrestrial laser scanning (TLS) data used for deriving quantitative structure models (QSMs) and fine-branching metrics. Data are grouped by study site: 2214, 2215, 2218, 2219, 2220, 2221 Each site folder contains three subfolders: 1_pointclouds/ Raw TLS point cloud data used to generate QSMs.Point clouds are voxelized to 0.015 m. 2_treegraph_qsm/ QSMs derived using the Treegraph library: Treegraph library v1.8 (Python v3.7) Source code for Treegraph: https://github.com/wanxinyang/treegraph Resulting QSMs from Treegraph include four files by default. (1) The .centre.ply files include the skeleton nodes of the reconstructed tree. (2) The .txt files provide the pointcloud data in text format. (3) The .mesh.ply files include fitted cylinders from the QSM models. (4) The .json files include all relevant information and are used for the tree_qsm.R script. More information can be found at Yang et al. 2024 (https://doi.org/10.1002/rse2.399). These QSMs represent the initial QSM reconstruction prior to additional correction/adjustment steps. 3_rtwig_adjusted_qsm/ This folder contains the adjusted/corrected QSM outputs produced by running tree_qsm.R.The script reads the Treegraph QSMs from 2_treegraph_qsm/, applies corrections using the rTwig library (https://cran.r-project.org/web/packages/rTwig/index.html), and derives tree-level metrics used in the manuscript. 2) Growth and vitality data (growth_and_vitality_data/) growth_data_intense.csv Retrospective growth dataset used for growth analyses and drought-related growth reduction estimates.The growth data originate from Schmied et al. (2023): https://doi.org/10.1016/j.foreco.2023.121305 Columns: site_idUnique site identifier (2214, 2215, 2218, 2219, 2220, 2221) siteSite name (e.g. Oberschwarzach, …) yearGrowth year tree_idUnique tree identifier (includes the site_id, e.g. 221401, 221402, …) rwRing width (mm) dbh_springDiameter at breast height (1.3 m) measured before the vegetation period of the respective year (cm) vitality_survey_intense.csv Vitality assessment dataset used in the manuscript (tree-level). Columns: site_idUnique site identifier (2214, 2215, 2218, 2219, 2220, 2221) tree_idUnique tree identifier (includes the site_id, e.g. 221401, 221402, …) dbhDiameter at breast height (1.3 m) measured at the end of 2022 (cm) crown_volumeCrown volume (m³) derived from crown hulls fitted to point clouds(see Bohnhorst et al. 2025: https://doi.org/10.1016/j.foreco.2025.123274) vitality_classBinary vitality classification (0/1) from Schmied et al. (2023). 0 = modest defoliation category 1 = pronounced defoliation category defoliation_01Estimated defoliation level in % for year 2022 defoliation_02Estimated defoliation level in % for year 2023 defoliation_03Estimated defoliation level in % for year 2024 defoliation_04Estimated defoliation level in % for year 2025 Script details growth.R Topic: Preparing growth data for subsequent analyses This script: estimates the potential population-level growth curve quantifies drought-induced growth reductions creates Figure S3 (Supplementary Material) tree_qsm.R Topic: Corrections of QSM models, visualisation and derivation of tree metrics This script: reads QSM models created with the Treegraph library adjusts/corrects QSM models using the rTwig library derives tree metrics from the QSM models calculates relative twig length generates visualisations for publication (Figures 2, 3, S2, S5) statistical_analyses.R Topic: Statistical analyses This script: develops the nonlinear model used for statistical analyses derives mortality and damage probabilities and generates predictions visualises model results (Figures 4, 5, S4, S6) checks model quality and performance uses data generated from growth.R and tree_qsm.R Software requirements Analyses were performed in R.The Bayesian modelling requires a functional Stan installation. Citation and reuse If you use this code or data, please cite: this Zenodo archive the associated manuscript Contact For questions, issues, or reproduction problems, please contact Gerhard Schmied ([e-mail]).