Reproducibility workflow and frozen input data for climate-risk prioritization of wild African Vigna

This Zenodo record contains the reproducibility workflow for the article “Climate-risk prioritization of wild African Vigna identifies complementary ex situ collection and in situ refugia priorities”. The record archives the R scripts, configuration files, documentation and frozen initial input data required to regenerate the analytical workflow. The workflow covers occurrence-data processing inputs, species distribution modelling, planning-unit feature construction, Collection Priority Index and rarity-weighted urgency calculations, prioritizr scenario optimization, quality-control checks and publication-output generation.

The archived input data include vigna_occurrences_cleaned.csv, a cleaned and taxonomically harmonized occurrence assembly derived from GBIF and Genesys records for wild African Vigna, the species selection input, and frozen third-party collectability inputs derived from OpenStreetMap/Geofabrik and ESA WorldCover. These inputs are included to freeze the exact analysis inputs used by the workflow and to support computational reproducibility.

Generated analysis outputs are intentionally not archived because of their size. Excluded outputs include SDM rasters, current and future suitability layers, planning-unit feature layers, CPI and rarity-weighted urgency outputs, prioritizr scenario solutions, selected planning units, maps, figures, publication tables and QC result tables. These outputs can be regenerated from the archived workflow and input data.

The record uses multiple/source-specific licenses. MIT applies only to original code and author-created documentation under code/. OpenStreetMap/Geofabrik-derived collectability files retain ODbL 1.0. ESA WorldCover files retain CC BY 4.0. The cleaned occurrence assembly is shared as a frozen reproducibility input and should be cited together with the original GBIF download DOI and Genesys provenance information documented in the bundle.

The record is organized as a structured reproducibility workflow. code/ contains the R scripts, configuration files and workflow documentation. input_data/ contains the frozen occurrence assembly, species-selection input and collectability inputs. reproduction/ contains instructions, data requirements, climate-input notes and a list of expected outputs that are intentionally not archived. SHA256 checksums and a deposition manifest are provided to document file integrity.

The workflow starts from a cleaned occurrence assembly of wild African Vigna records derived from GBIF and Genesys, together with species-selection inputs and frozen collectability layers. The R workflow then reconstructs the analytical sequence used in the associated publication: species distribution modelling inputs and consensus objects, 10 km planning-unit features, climate-risk and persistence metrics, collection-priority and rarity-weighted urgency indices, prioritizr conservation-planning scenarios, scenario comparison and quality-control summaries. Generated outputs are written locally when the workflow is rerun and are not included in this archived record.

The occurrence file vigna_occurrences_cleaned.csv is a frozen cleaned assembly of GBIF and Genesys records for wild African Vigna. It is included as the exact occurrence input used by the modelling and prioritization workflow. This processed assembly does not replace the original GBIF and Genesys sources; users should cite the original sources using the GBIF download DOI and Genesys provenance information documented in the bundle.