Training and intercalibration reduce observer-induced variability in forest vegetation surveys

Data supporting and allowing reproducibility for the article Cervellini, M. et al. (2026) «Training and intercalibration reduce observer-induced variability in forest vegetation surveys», Vegetation Ecology and Diversity.

File structure

Data are provided as a .zip file to preserve the folder structure. The subfolders 2023 and 2025 separate the two training events. The original data structure has been preserved, and all the data handling has been done through the R script. The 2023 data is collected in a single file, while the 2025 data has individual files for each observer. The 25_02 folder collects the pre-training data, while the 27_02 is post-training. Files in the 25_02 folder have an additional "Plot" sheet that has not been used.

Field description

2023 sheet

• Anno:  sampling year
• Id Plot: unique identifier of the site-observer combination
• Numero Plot: code identifying the observer team and whether it is before or after training; 32–35 the four pre-training teams, 36–39 the four post-training teams
• Subplot: number of the plot (1–4)
• In/Out: whether the survey is within a fence or not. Mandatory field for ICP Forests data, not used here (always 1)
• Primavera/Estate: whether the survey is in Spring or Summer (always 1, Spring)
• Codice Strato: layer code; 1. tree (woody + pertinent lianas and climbers ) h > 5 m, 2. shrub (woody + pertinent lianas and climbers ) 0.5 < h ≤ 5, 3. herb(however herbaceous and ferns; woody only if h ≤ 0.5 m) h ≤ 0.5 m, 4. moss (terricolous bryophytes and t-lichens) on mineral, organic soil
• Genere e Specie Pignatti: genus and species name according to "Flora d'Italia", Pignatti S. 1982
• Copertura Braun-Blanquet: cover according to the Braun-Blanquet cover-abundance scale (+, r, 1, 2, 3, 4, 5)
• Copertura %: percent cover derived from the median values of the Braun-Blanquet scale (0.01, 0.5, 3, 15, 37.5, 62.5, 87.5)
• Note: additional notes

2025 sheets

• Plot.code: identifier composed of the observer letter ID (a–q) and the year (2025)
• SU: number of the plot (1–8)
• Layer: vegetation layer; T. tree, S. shrub, H. herb
• species (Genus + species): genus and species name
• Cover(perc.): percent cover (0.001–100)
• Notes: additional notes

Code and notebook

Script and rendered notebook are available at the linked GitHub/Zenodo repository at the end of this page.