GBIF Animal Occurrence Records - CC0 Georeferenced

GBIF-mediated animal occurrences as clean, per-record biodiversity data - one record per occurrence the publisher released under CC0 1.0, each carrying decimal latitude/longitude. Taxonomy, institution codes, locality and dates, with collector and identifier person names structurally excluded.

What one record represents

The source is GBIF (the Global Biodiversity Information Facility) occurrence search API, filtered to the animal kingdom (kingdomKey=1) with coordinates (hasCoordinate=true), released under CC0. Each record is one occurrence: its GBIF id, publishing dataset key, basis of record, catalogue/occurrence identifiers and institution/collection codes; the full taxonomy (kingdom through species, plus the accepted scientific name, taxon rank and keys); locality (country, country code, state/province) with decimal latitude/longitude, coordinate uncertainty, elevation and depth; the event date (year, month, day and verbatim date); and type status, establishment means and individual count where recorded.

Sample output

Real rows from a live run of this actor (first 5 rows, selected columns).

One full record from the same run, exactly as delivered:

{
  "gbif_id": "1826325281",
  "dataset_key": "7e380070-f762-11e1-a439-00145eb45e9a",
  "basis_of_record": "PRESERVED_SPECIMEN",
  "occurrence_id": "6ec89d63-5cc1-4b7c-b020-addf780ffbf7",
  "catalog_number": "BMNH(E)1900813",
  "institution_code": "NHMUK",
  "collection_code": "BMNH(E)",
  "publishing_org_key": "19456090-b49a-11d8-abeb-b8a03c50a862",
  "kingdom": "Animalia",
  "phylum": "Arthropoda",
  "class_name": "Insecta",
  "order_name": "Lepidoptera",
  "family": "Geometridae",
  "genus": "Agriopis",
  "species": "Agriopis marginaria",
  "scientific_name": "Agriopis marginaria (Fabricius)",
  "accepted_scientific_name": "Agriopis marginaria (Fabricius)",
  "taxon_rank": "SPECIES",
  "taxon_key": 5769068,
  "kingdom_key": 1,
  "type_status": null,
  "establishment_means": null,
  "individual_count": 1,
  "country": "Ireland",
  "country_code": "IE",
  "state_province": "Wicklow",
  "decimal_latitude": 52.9323,
  "decimal_longitude": -6.2311,
  "coordinate_uncertainty_m": null,
  "elevation": null,
  "depth": null,
  "year": 2026,
  "month": 1,
  "day": null,
  "event_date": "2026-01",
  "license": "CC0 1.0",
  "record_id": "1826325281",
  "source": "GBIF (Global Biodiversity Information Facility)",
  "source_dataset": "occurrence/search?kingdomKey=1&hasCoordinate=true&license=CC0_1_0",
  "licence": "GBIF-mediated occurrence data. This cell delivers ONLY records the publisher released under CC0 1.0 (public-domain dedication; license=CC0_1_0); no rights are reserved on the data values. Use is governed by the GBIF Terms of Use (https://www.gbif.org/terms). GBIF asks that GBIF.org and the publishing datasets be cited; dataset_key identifies the publisher.",
  "attribution": "GBIF.org and the publishing institutions (per dataset_key)",
  "caveat": "Georeferenced animal-kingdom (kingdomKey=1) occurrence records released under CC0 1.0 (about 226.6 million match the filter live), each carrying decimal latitude/longitude. GBIF caps offset paging at 100,000, so this cell delivers the first 100,000-record accessible slice; narrow downstream by year, country or dataset to reach records past the cap. Collector and identifier person names (recordedBy, identifiedBy and every *By field) are structurally excluded.",
  "observed_at": "2026-09-25T17:23:10Z"
}

Coverage and volume

About 226.6 million georeferenced CC0 animal occurrences match the filter live - 226,606,952 measured at build time from the API's match count. Every record carries coordinates.

Sol's Wave-3 index put this door at about 226,658,291 matches; measured live at build time 226,606,952 match under CC0 with coordinates - the live figure is what this listing quotes, delivered as a 100,000-record accessible slice.

GBIF caps offset paging at 100,000 (a request past that returns HTTP 400), so this cell delivers the first 100,000-record accessible slice of the match set. To reach records beyond the cap, narrow the query downstream by a facet the API partitions on - year, country or datasetKey - each of which yields its own sub-100,000 window; the baked filters (kingdom, coordinates and the CC0 licence) always apply. The Actor pages 300 occurrences at a time and stops as soon as your Maximum records cap or the 100,000 ceiling is met.

Licence and attribution

This cell delivers only occurrences the publisher released under CC0 1.0 - a public-domain dedication with no rights reserved on the data values. The licence filter (license=CC0_1_0) is baked into every query, not a buyer option, and a per-record assertion rejects any occurrence whose licence is not CC0 (verified with a planted-row test). The full notice travels on every record:

GBIF-mediated occurrence data. This cell delivers ONLY records the publisher released under CC0 1.0 (public-domain dedication; license=CC0_1_0); no rights are reserved on the data values. Use is governed by the GBIF Terms of Use (https://www.gbif.org/terms). GBIF asks that GBIF.org and the publishing datasets be cited; dataset_key identifies the publisher.

GBIF asks that GBIF.org and the contributing datasets be cited; dataset_key on every record identifies the publisher so you can attribute or trace provenance per row.

Person-data policy

GBIF occurrences carry collector and identifier person names (recordedBy, identifiedBy, and every *By / *ByIDs field). This Actor structurally excludes all of them: they are never copied into a record. Only institution and collection codes, taxonomy, locality, coordinates and dates are emitted, and a per-record assertion rejects any collector/observer/person field (verified with a planted-field test). No natural-person data is processed.

Interpretation caveat

Georeferenced animal-kingdom (kingdomKey=1) occurrence records released under CC0 1.0 (about 226.6 million match the filter live), each carrying decimal latitude/longitude. GBIF caps offset paging at 100,000, so this cell delivers the first 100,000-record accessible slice; narrow downstream by year, country or dataset to reach records past the cap. Collector and identifier person names (recordedBy, identifiedBy and every *By field) are structurally excluded.

Values are reproduced verbatim from the API; the Actor never rewrites a field. Coordinates are the publisher's decimal degrees; because hasCoordinate=true is baked, every record has them. individual_count and type_status are present only where the publisher recorded them.

Data quality and freshness

Numeric fields (coordinates, uncertainty, elevation, depth, year/month/day, taxon keys, individual count) are delivered as real numbers; the licence label is normalised to CC0 1.0. Delivery is keyed on the stable GBIF id, so the dataset is safe to diff, deduplicate or upsert. Every run re-reads the live API, so the data is as fresh as GBIF's index, and each record's observed_at stamp dates the snapshot. The run's RUN_RECEIPT records the live match count alongside how many records were delivered and charged, and the 100,000 hard cap.

Provenance and compliance

Every run reads api.gbif.org/robots.txt at runtime; the gate result (URL, status, byte length, SHA-256 of the policy) is written to the run's RUN_RECEIPT, and the occurrence-search path is confirmed crawlable before any data request. The API is keyless. The Actor never bypasses a block or fetches through a mirror.

Inputs

• Maximum records (maxRecords) - hard cap on occurrence records delivered and billed (also ceilinged at the 100,000 accessible slice). You are billed only for records actually delivered.

Output

Records land in the Actor's default dataset and export as JSON, CSV, Excel or via the Apify API. A tabular overview view surfaces the GBIF id, scientific name, basis of record, country code, coordinates, year and institution code.

Fields in detail

The record leads with gbif_id and dataset_key, then the basis of record and identifiers (occurrence_id, catalog_number, institution_code, collection_code, publishing_org_key); the taxonomy block (kingdom, phylum, class_name, order_name, family, genus, species, scientific_name, accepted_scientific_name, taxon_rank, taxon_key, kingdom_key); type/ecology (type_status, establishment_means, individual_count); geography and time (country, country_code, state_province, decimal_latitude, decimal_longitude, coordinate_uncertainty_m, elevation, depth, year, month, day, event_date); and the normalised license. The provenance block closes every record.

The third GBIF cell

This completes the fleet's GBIF trio. The three cells draw the same door and the same recipe, and stay distinct by the baked occurrence filter: the preserved-specimen cell is museum/herbarium specimens (basis of record), the plant cell is the plant kingdom georeferenced, and this one is the animal kingdom georeferenced. Each is a windowed, licence-clean feed of its slice; together they cover the CC0 georeferenced animals, the CC0 georeferenced plants, and the CC0 preserved specimens without overlapping in scope.

Typical uses

Nature-risk, conservation and ecology teams use this cell to screen animal presence - a windowed, licence-clean occurrence feed with coordinates, ready to drop into a species-distribution model, a range map, or a site-screening query, without touching GBIF's Darwin Core archives or download API. Because the taxonomy is fully expanded and coordinates are real numbers, the records feed straight into spatial joins against exposure or land-use data. To assemble a national or multi-year panel that exceeds 100,000 records, run the cell once per year, country or datasetKey window and concatenate - every window inherits the baked CC0, kingdom and coordinate filters. A scheduled run keeps a downstream table current with GBIF's index, and the stable GBIF id makes the dataset safe to upsert.

Scaling and limits

Set Maximum records low to sample cheaply or high to pull toward the 100,000 accessible slice. The Actor pages the API 300 records at a time and delivers incrementally, so memory stays flat, and you are billed only for what is delivered. Because the source holds far more than 100,000 matches, treat this cell as a windowed feed and concatenate windows for a larger panel; re-running picks up GBIF's newly indexed occurrences, and each record's observed_at stamp dates the snapshot.

Sibling Actors

It sits beside the fleet's GBIF preserved-specimen and plant-occurrence cells (the same door, different baked filters) and completes the fleet's CC0 GBIF coverage. It shares its engineering - the runtime robots gate, offset paging with the licence filter baked in, push-then-charge billing and verbatim-value discipline - with the fleet's other GBIF cells, from which it differs only in the baked occurrence filter (animal vs plant vs preserved specimen).