Code and data for paper "Scale dependence of avian functional rarity reveals mismatches between global and local conservation priorities"

1) What’s inside

• FREE_BIRD.Rproj – open this to work within the project root in RStudio.

• analysis/ – step-by-step analysis scripts:

  • A_prepare_data_and_compute_rarities.R

  • B_Vizualize_species_metrics.R

  • C_map_multiscale-prioritisation.R

• R/ – helper functions used by the analysis (e.g., function_compute_pool_distinctiveness.R, function_diversity_optimisation.R, find_neighbors.R, plotting/transformation utilities).

• data/SpatialData/ – spatial layers and precomputed helper tables (global grid, land/ocean/coastline shapes, ecoregions, protected area coverage, neighbor lists, etc.). Several are large (hundreds of MB to a few GB). Examples:

  • AllSpeciesBirdLifeMaps2019.csv

  • neighbor tables (e.g., cell_neighbors_within_30k.csv, cell_neighbors_within_distance.csv)

  • Natural Earth and other shape files (ne_110m_*, ne_50m_coastline, world_ecoregions, Behrmann meter grids), each with their own small READMEs/VERSION notes.

• data/TraitData/ – AVONET tables (AVONET_Raw_Data.csv, species list, sources, duplicates, etc.).

• figures_tables/ – rendered figures and tables used in the paper; includes the folder supp_FigureWithoutSEAbirds/ with sensitivity figures that exclude seabirds.

• outputs/ – created when you run the scripts; contains intermediate objects and exported rasters/maps (will be populated by the analysis).

The archived deposit is released under CC-BY-4.0; please credit the authors when reusing code or derived data.

2) Requirements

• R (RStudio recommended). The project is R-based.

• Disk & RAM. Several inputs are multi-GB (e.g., neighbor tables); ensure ample disk space (tens of GB) and sufficient memory for large table joins and raster operations.

• Packages. The scripts will load typical spatial/data packages (e.g., data.table/readr, sf, terra/raster, dplyr, ggplot2, etc.). If any package is missing, install on first run.

3) Quick start

1. Download & unzip FREE_BIRD_repo.zip.

2. Open FREE_BIRD.Rproj in RStudio (your working directory will be the project root).

3. Run the scripts in order (A → B → C) from the analysis/ folder:

   • A_prepare_data_and_compute_rarities.R

     • Reads AVONET traits (data/TraitData/) and species distributions / spatial scaffolding (data/SpatialData/).

     • Computes functional distinctiveness/restrictiveness and composite rarity across the defined spatial grains; writes cleaned/derived objects to outputs/. 

   • B_Vizualize_species_metrics.R

     • Produces species-level summaries/diagnostics and figures (e.g., PCA of distinctiveness axes, scale comparisons); saves figures into figures_tables/ and intermediate summaries into outputs/.

   • C_map_multiscale-prioritisation.R

     • Generates gridded maps of richness/rarity/prioritization at multiple scales and writes GeoTIFF/PNG outputs to outputs/ and figures_tables/.

Tip: run each script from the project root so relative paths (e.g., data/SpatialData/...) resolve correctly.

4) Inputs you should know about

• Species distributions: AllSpeciesBirdLifeMaps2019.csv (BirdLife consolidated presence by grid cell).

• Spatial scaffolding: Behrmann meter grid shapefiles (reference_grid/, worldBehrmannMetergrid_WGS84/), land/ocean/coastline layers (ne_110m_*, ne_50m_coastline), ecoregions, hotspots; each subfolder ships its own small README/VERSION or metadata file.

• Trait data: AVONET CSVs under data/TraitData/. 

• Neighbor tables: large CSVs used to compute within-distance neighborhood metrics (e.g., 30 km, 300 km). Expect long read times. 

5) Expected outputs

• Recreated figures and tables will appear under figures_tables/ (e.g., global maps, PCA plots; the repository already includes a set of rendered figures, including those without seabirds for sensitivity). 

• Intermediate data products and final rasters/CSV summaries will be written under outputs/ by scripts B and C. (Folder created on first run.) 

6) Reproducing the paper figures

• After running A and B, you should recover the species-level plots (PCA, rarity components).

• After C, you should recover the multiscale global maps used in the main text and supplement; check figures_tables/ against the provided supp_FigureWithoutSEAbirds/ examples to verify consistency. 

7) Notes, licensing, and attribution

• The Zenodo record is Version 1 (published 3 Nov 2025) and is licensed CC-BY-4.0; cite the deposit DOI in any reuse. 

• Some third-party layers in data/SpatialData/ include their own README/VERSION files; please respect their original terms when redistributing derived products.