Species range shifts often speed ahead of their modeled climatic niches

This repository contains the code required to reproduce the analyses presented in the study:

Oliveira et al. (2026). Species range shifts often speed ahead of their modeled climatic niches. PNAS.

The study evaluates how well climate-based species distribution models (SDMs) predict the direction and rate of observed species range shifts.

Repository Structure

├── R/                               # R scripts for data processing, modeling, and analysis

│   ├── 0_supplementary/              # Scripts used for supplementary analyses
│   │   ├── 1_get_sps_for_supp.R      # Select random 10% of terrestrial and marine species
│   │   ├── 2_get_bioclimatics_sps.R  # Extract bioclimatic variables for each species
│   │   ├── 2_job_bioclimatics_sps.R  # Submit previous script to the HPC
│   │   ├── 3_run_sdm.R               # Run SDMs for supplementary species
│   │   ├── 3_job_sdm.R               # Submit SDM script to the HPC
│   │   ├── 4_gather_sdm_CV.R         # Gather model cross-validation results
│   │   ├── 5_get_shifts_edge_method.R# Calculate species range shifts from SDMs
│   │   ├── 5_job_shifts.R            # Submit range shift script to the HPC
│   │   ├── 10_merge_all.qmd          # Merge modeled and documented range shifts
│   │   └── 10_analyses.qmd           # Analyses comparing documented and modeled shifts

│   ├── 1_select_species/             # Select species for downstream analyses
│   │   └── 1_get_species_from_SDMs.R

│   ├── 2_bioclimatic_data/           # Calculation of bioclimatic variables

│   │   ├── CHELSA/                   # Climate data for terrestrial species
│   │   │   ├── 1_CHELSA_download.R           # Download climatic variables from CHELSAcruts
│   │   │   ├── 2_CHELSA_bioclimatic_proj.R   # Calculate temperature and precipitation summaries
│   │   │   ├── 2_CHELSA_bioclimatic_proj.sh  # Submit previous script to the HPC
│   │   │   └── 2_loop_years_proj.sh          # Loop batch submission across years

│   │   └── ORAS/                     # Climate data for marine species
│   │       ├── 1_download_oras5.py        # Download ORAS5 data
│   │       ├── 2_unzip_and_save.R        # Unzip and store downloaded files
│   │       ├── 2_ORAS_job_unzip_save.sh  # Submit unzip script to the HPC
│   │       ├── 3_ORAS_bioclimatics_proj.R# Calculate SST statistics
│   │       └── 3_ORAS_job_proj.sh        # Submit previous script to the HPC

│   ├── 2_2_bioclimatic_SA/           # Bioclimatic variables for study areas
│   │   ├── get_bios_SA.R             # Calculate yearly bioclimatic variables for each study area
│   │   └── job_bios_SA.R             # Submit previous script to the HPC

│   ├── 3_gbif_data/
│   │   └── 1_Get_GBIF_data.R         # Download and filter GBIF occurrence records

│   ├── 4_bioclimatics_sps/
│   │   ├── get_bioclimatics_sps.R    # Extract bioclimatic variables for GBIF records
│   │   └── job_bioclimatics_sps.R    # Submit previous script to the HPC

│   ├── 5_sdms/
│   │   ├── 1_run_sdm.R               # Run species distribution models
│   │   ├── 1_job_sdm.R               # Submit SDM jobs to the HPC
│   │   └── 2_gather_sdm_CV.R         # Gather SDM cross-validation results

│   ├── 6_range_shifts/
│   │   ├── 1_get_shifts_edge_method.R# Calculate species range shifts from SDMs
│   │   ├── 1_job_shifts.R            # Submit range shift script to the HPC
│   │   └── 2_gather_shifts_edge_method.R # Compile range shifts across species

│   ├── 9_connectivity/
│   │   ├── 1_get_connectivity_SA.R   # Calculate connectivity for each study area
│   │   ├── 1_job_connectivity_SA.R   # Submit previous script to the HPC
│   │   ├── 2_get_connectivity_sps.R  # Calculate species-specific connectivity
│   │   └── 2_job_connectivity_sps.R  # Submit previous script to the HPC

│   └── 10_analyses/
│       ├── 1_merge_all.qmd           # Compile documented and modeled range shifts
│       ├── 2_data_exploration.qmd    # Exploratory data analysis
│       └── 3_SDM_models.qmd          # Multivariate statistical analyses


├── Data/
│   └── Output/
│       ├── Bioshifts_merge_Exposure_all.csv
│       │   # Dataset containing merged documented and modeled range shifts
│       └── Bioshifts_merge_Exposure_all_supp.csv
│           # Dataset used in supplementary analyses

├── renv/                             # Project-specific R package library
├── renv.lock                         # Reproducible R package environment
└── README.md                         # Project documentation

Workflow

1. Select species with documented range shifts in the BioShifts and CoRE databases.

2. Download species occurrence data from GBIF

3. Download environmental data

   • Terrestrial species: CHELSAcruts climate datasets

   • Marine species: ORAS oceanographic datasets

4. Fit ensemble species distribution models (SDMs) for each species.

5. Estimate modeled range shifts using SDM projections across time.

6. Compare modeled and documented range shifts to assess agreement in direction and magnitude.

Reproducibility

This project uses the renv package to ensure reproducibility of the R environment.

To recreate the computational environment:

install.packages("renv") 

renv::restore()

This will install all package versions recorded in renv.lock.

Data Availability

• Species occurrences: GBIF (https://www.gbif.org)

• Documented range shifts: BioShifts and CoRE databases

• Environmental data:

  • CHELSAcruts for terrestrial species (http://chelsa-climate.org)

  • ORAS5 for marine species (https://cds.climate.copernicus.eu/datasets/reanalysis-oras5?tab=overview)

⚠️ Raw data may require downloading separately due to size or licensing.

Citation

If you use this code or build upon this work, please cite:

Oliveira, B. F. et al. (2026). Species range shifts often speed ahead of their modeled climatic niches. Proceedings of the National Academy of Sciences (PNAS).