How we read the records

This site is built on occurrence records — dated, located observations of a species. That is a different thing from a range map, and the difference matters.

Where the data comes from

Every count here derives from GBIF, which pools records from museums, national surveys and citizen-science projects such as eBird and iNaturalist. A record says "someone identified this species at this place on this date". It does not say the animal breeds there, is common there, or is still there.

What we exclude

We drop fossil specimens and records of living captive animals, because a fossil or a zoo enclosure is not evidence that an animal lives somewhere. We keep only records with coordinates.

How status is decided

Each animal gets a status from the shape of its record history, not from an authority:

• resident recorded across many years and spread through the year.
• seasonal recorded repeatedly, but concentrated in part of the year.
• vagrant very few records — likely a stray, an escape, or a one-off.
• historical nothing recent; the presence is in the past.

The biases you should assume

Recording effort is wildly uneven. Birds dominate because birdwatchers submit vast numbers of checklists; insects and marine life are under-recorded nearly everywhere. Rich, well-surveyed countries look biodiverse partly because they are well surveyed. A high count means "well recorded", not "abundant".

Credits

Occurrence data: GBIF.org, CC-BY / CC0 / CC-BY-NC per contributing dataset. Photographs: Wikimedia Commons, each credited to its photographer with its licence on the page where it appears.