GBIF

Primary source
Global Biodiversity Information Facility
An intergovernmental data network, funded by national governments, that pools biodiversity records from museums, herbaria, research projects, and citizen-science platforms into one free archive of where species have actually been recorded. It also publishes the GBIF Backbone Taxonomy, the single name list every one of those records is matched against, which is where its accepted names and synonyms come from.
What you'll find here
Occurrence records
Observation/specimen evidence
Distribution context
Dataset citations
A second opinion on the accepted name
Specific data we draw from it: occurrence count, country/state evidence, dataset DOI, basis of record, backbone accepted name + taxonomic status.
Good to know
Occurrence confidence. Separate specimen, observation, literature, and geospatial-quality issues before summarizing range.
Media licensing. Treat occurrence images separately because media licenses can differ from occurrence-data licenses.
The backbone is its own treatment. The GBIF backbone is assembled from many contributed checklists and is not Kew’s. It can keep a familiar name as accepted for years after Kew has moved the species to another genus, so a difference between the two is a real disagreement between authorities to report, not an error to quietly correct.
How to cite this source
Cite GBIF downloads or dataset pages, including DOI/download key when available.
License
Occurrence records and media can carry different terms; check the dataset and multimedia extension license.
Caution
Occurrence records are evidence, not garden suitability recommendations. Review geospatial flags before using them. The backbone taxonomy is a name-matching index rather than a taxonomic opinion of record, so cite it alongside Kew where the two differ and never as the sole authority on an accepted name.