Before you cite anything
Reading the data
Every number in this site comes from GBIF, the Global Biodiversity Information Facility. Knowing what those numbers are actually counting is the difference between a good ecography and a confident wrong one.
What GBIF is
An international network that publishes species occurrence records — more than three billion of them. A record says: this species was here, on this date, according to this person or institution.
The records come from natural history museums digitizing century-old drawers, from university field surveys, from government monitoring programs, from bird counts, and — in enormous and growing numbers — from members of the public photographing things with their phones.
What stays private
- No account, no login, no analytics, no third-party trackers.
- Your circle, your collected species, and your writing live in this browser's local storage and are sent nowhere until you press export.
- If you use Use my location, the coordinate is used to center the map and query GBIF, stored on this device, and shared with no one.
- Outbound requests go only to GBIF (species records), Wikipedia (species descriptions), OpenStreetMap (map tiles and town search), and Google Fonts. None of them receive your writing.
- Searching for a town sends that search text to OpenStreetMap's geocoder. It is the only thing you type that ever leaves the device.
- Clearing your browser data clears your work. Export before you finish.
Five things a record count cannot tell you
Read these before you write a sentence that starts "the data shows".
1. It is not a population
A thousand records of a robin and four of a salamander does not mean there are more robins. It means robins are loud, daytime, and easy to photograph, and salamanders are under a log at night.
2. Sampling is wildly uneven
Records cluster along roads, trails, campuses, and coastlines, and in wealthier countries and neighborhoods. The map of records is partly a map of who had leisure time, transport, and a smartphone.
3. Absence is ambiguous
Zero records can mean locally extinct, never surveyed, never digitized, or recorded under a different scientific name. Four very different stories, one identical number.
4. The past is thin, and getting thinner backwards
Almost every place shows a steep rise in records after about 1990. That is databases and phone apps arriving, not life arriving. Old records are often dated to a year or a decade after the fact.
5. Coordinates carry uncertainty
Many records list an accuracy radius in meters. Some were georeferenced decades later from a written description like "6 miles north of town". A dot on a map looks more certain than it is.
And one thing it is very good for
Asking better questions. A surprising absence, an odd cluster, a season that does not match what you have seen with your own eyes — those are the starting points for real inquiry.
Citing this properly
Your exported ecography includes these automatically, but you should understand them.
The search you ran
GBIF asks that you cite the search itself, with the date, because the database changes daily:
GBIF.org (YYYY-MM-DD) GBIF Occurrence Search. https://www.gbif.org/occurrence/search
The Wikipedia descriptions
Each species card pulls the opening of its Wikipedia article so you can find out what the thing actually is. Wikipedia text is licensed CC BY-SA 4.0, and your exported ecography credits it automatically. Wikipedia is a fine place to start and a poor place to stop — if you use a fact from it in your writing, follow its footnote to the original source.
The individual records
Each record belongs to the institution or person who published it, and carries its own license — often CC0, CC BY, or CC BY-NC. Photographs are licensed separately from the record. CC BY-NC images cannot be used commercially, which includes some school publishing. Every photo in this site shows its photographer and license.
For instructors
A first-week freewrite
Assign one circle around campus or the student's hometown, two or three collected species, and all three writing sections. Roughly 45–60 minutes. Students submit the exported HTML.
What to grade: specificity. Whether the reader can picture the place, and whether the species are described rather than listed.
A data-literacy discussion
Have two students draw circles of the same size — one over a city park, one over farmland or forest — and compare record counts.
The point: the difference is mostly observers. Recognizing sampling bias transfers far beyond ecology.
Fieldwork that closes the loop
Students draw a small circle around somewhere they can physically walk, list five species the data predicts, then go and check.
The payoff: both matches and misses are worth writing about, and the third section is built for exactly that.
Adapting the site
Starting points, group definitions, and the writing prompts all live in
data/starting-points.json. Edit that file to change the assignment — no
code changes needed.
The whole site is static: no build step, no server, no keys.