Darwin Core

Darwin Core (frequently abbreviated DwC) is a data standard for sharing biodiversity data — records of the occurrence of organisms, and the specimens, observations, and events that document them. It was ratified as a formal standard by the Biodiversity Information Standards organization (TDWG, formerly the Taxonomic Databases Working Group) on 9 October 2009, and its List of Terms has been maintained on an ongoing basis since by the TDWG Darwin Core Maintenance Group rather than through numbered major-version releases the way some other metadata schemas are.

Core classes and terms

Darwin Core defines a small set of classes that a dataset’s fields are mapped onto, plus the individual terms within each. The classes most commonly used for occurrence and specimen data include:

• Occurrence — a documented presence (or absence) of an organism at a location and time; the central class for species-occurrence records.
• Taxon — the scientific name and classification applied to a record (dwc:scientificName, dwc:taxonRank, dwc:kingdom, and related terms).
• Event — the action or set of circumstances during which an occurrence was recorded (a collecting trip, a camera-trap deployment, a single observation), carrying dwc:eventDate and other temporal terms.
• Location — the spatial region or named place associated with an event or occurrence (dwc:decimalLatitude, dwc:decimalLongitude, dwc:coordinateUncertaintyInMeters, dwc:country).
• MaterialEntity — a physical or preserved thing such as a specimen or sample that can itself be tracked and cited.
• Identification — a taxonomic determination applied to an occurrence or material entity, including who made it and when.

Record-level terms (identifiers, licensing, rights, basisOfRecord) plus Agent, GeologicalContext, MeasurementOrFact, and ResourceRelationship classes round out the full term list — see the normative reference below for the complete set.

Real-world use: GBIF and biodiversity data aggregation

The Global Biodiversity Information Facility (GBIF), the largest aggregator of species-occurrence data, publishes the majority of the datasets on its network in the Darwin Core Archive (DwC-A) format: a self-contained package of one or more Darwin Core-mapped data files plus a machine-readable descriptor (meta.xml) that maps each column to its Darwin Core term, typically alongside Ecological Metadata Language (EML) dataset-level metadata. Data publishers commonly build these archives through GBIF’s Integrated Publishing Toolkit (IPT), which handles the term-mapping and packaging step. Other biodiversity data infrastructures — including national GBIF nodes and the Ocean Biodiversity Information System (OBIS) — likewise expect Darwin Core-formatted, DwC-A-packaged submissions, since a shared term set is what lets occurrence and specimen records from thousands of independent collections, museums, and citizen-science projects be combined into a single searchable index.

A discipline-specific example of metadata schema selection

Darwin Core is a useful illustration of the general principle covered in CASRAI’s guide to choosing a metadata schema for a dataset: the right schema is the one a dataset’s target repository or aggregator actually expects, not a generic default. A biodiversity-occurrence dataset destined for GBIF or OBIS should be described in Darwin Core rather than a general-purpose schema like the DataCite metadata schema, which most repositories use for the DOI-registration metadata that sits alongside — not instead of — the Darwin Core-formatted data itself.

References

• TDWG (Biodiversity Information Standards), ‘Darwin Core’ — normative term list and quick reference guide (dwc.tdwg.org).
• TDWG standards register entry for Darwin Core, ratified 2009-10-09 (tdwg.org/standards/dwc/).
• GBIF, ‘Darwin Core Archives — How-to Guide’, GBIF IPT User Manual (ipt.gbif.org/manual/en/ipt/latest/dwca-guide).