Sourcing & Editorial Standards
Our wildlife guides, comparisons and rescue guides are researched, sourced and reviewed by the WARN Research & Conservation Team. Here is exactly where the facts come from — and how you can check our work.
By the WARN Research & Conservation Team · Last updated
In brief
WARN’s species facts come first from the IUCN Red List and CITES, then peer-reviewed science; every guide lists its sources, dates its figures, and carries a visible “Last updated” date. We revise pages when the underlying science changes and correct errors openly.
Our Source Hierarchy
- 1
Primary conservation authorities
The IUCN Red List of Threatened Species (for status, population trend and range) and CITES (for trade listings) are our first reference for every species. Where we state a status or a population figure, it is drawn from a published assessment and dated to its assessment year.
- 2
Peer-reviewed and institutional science
For biology, behaviour, taxonomy and ecology we rely on peer-reviewed literature and established scientific databases maintained by universities, natural-history museums and government wildlife agencies.
- 3
Reference encyclopaedias
General-reference works are used only to corroborate uncontested facts and to cross-check figures — never as the sole source for a conservation claim.
How We Work
Every claim is sourced
We do not publish population numbers, conservation statuses or biological “facts” we cannot trace to a named authority. Each guide lists its sources, and contested or time-sensitive figures carry their assessment year in the text.
We show our working
Guides display a visible “Last updated” date tied to the page’s modification date, so readers and AI assistants can see how current the information is. Comparison pages lead with a plain-language answer and a side-by-side table so the underlying facts are easy to check.
We update when the science does
When the IUCN publishes a new assessment or a widely reported figure changes, we revise the affected guides and refresh the “Last updated” date. Priority goes to the species we fund work for and the highest-traffic pages.
We separate fact from appeal
Educational content and fundraising are kept distinct: species facts stand on their own sources, and any call to support our work is clearly marked as such.
We correct openly
If we get something wrong, we fix it and update the modification date. Spotted an error? Email us and we will review it against the sources.
Who Writes and Reviews
Our guides are produced by the WARN Research & Conservation Team — the researchers and editors responsible for the accuracy of our educational content. Every guide is checked against the source hierarchy above before publication and revisited when new assessments are released.
World Animal Rescue Network is a registered global not-for-profit animal welfare organisation in England & Wales, funding partner-led animal rescue worldwide. Questions about a specific figure or source? Contact us and we will point you to the assessment behind it.
The Sources We Cite
Grouped by what each source is actually authoritative for. Citing a conservation database for husbandry advice, or a health agency for taxonomy, would be misuse — so we do not do it.
Conservation status, population and range
- IUCN Red List of Threatened Species — Threat category, population estimate and trend, range, assessment year. Our first reference for every species guide.
- CITES Appendices — International trade listings — whether a species is Appendix I, II or III, and what that permits.
- UNEP-WCMC Protected Planet — Protected-area coverage where a guide discusses habitat protection.
Taxonomy, occurrence and biology
- GBIF — Global Biodiversity Information Facility — Accepted scientific names, taxonomic hierarchy and verified occurrence records.
- Peer-reviewed literature — Behaviour, physiology, ecology and any contested claim. Preferred over secondary reporting in all cases.
- Natural history museum and university collections — Morphology, distribution history and taxonomic revisions.
Animal health, zoonoses and welfare practice
- World Health Organization — Rabies control, zoonotic disease and the evidence base for humane dog population management (CNVR).
- World Organisation for Animal Health (WOAH) — Animal health standards and welfare codes.
- FAO — Food and Agriculture Organization — Working animals, livestock welfare and pollination services.
Corroboration only
- General reference encyclopaedias — Used to cross-check uncontested background detail and to find primary citations. Never the sole source for a conservation, population or welfare claim.
How a Guide Gets Fact-Checked
Every guide passes through the same six steps before it is published.
- 1
Scope the question
We start from what a reader is actually asking — the query, not the species. That decides whether a page is a species guide, a comparison, a rescue guide or a short answer.
- 2
Pull the primary assessment
IUCN status, population trend and assessment year, plus the CITES listing where trade is relevant. These values are recorded with their year so nothing silently ages.
- 3
Build the biology from science
Taxonomy is checked against GBIF; behaviour, physiology and ecology come from peer-reviewed literature or institutional sources, not from other popular animal sites.
- 4
Write the answer first
Every guide opens with a direct one-sentence answer, then key takeaways, then the detail. If the opening sentence cannot be supported by the sources listed, the page does not publish.
- 5
Check the claims line by line
A second pass verifies every number, date, status and superlative against the cited source. Claims that cannot be traced are cut rather than softened.
- 6
Date it and publish
The page carries a visible "Last updated" date matching its structured-data dateModified, and lists its sources with working links.
When We Update, and When We Don't
| Trigger | What happens |
|---|---|
| A new IUCN assessment is published | The affected guides are revised and re-dated, prioritising species WARN funds work for. |
| A CITES listing changes | Trade sections and any status badge are corrected at the next review. |
| A reader reports an error | Reviewed against the sources; if upheld, corrected and re-dated, normally within five working days. |
| No change in the underlying science | Pages are left alone. We do not re-date a page to fake freshness — a "Last updated" stamp with no change behind it is a false signal. |
AI, Automation and Accountability
AI tools may be used for drafting support, structure and copy-editing. No factual claim is published on the basis of an AI output.
- Every conservation status, population figure, date and biological claim is verified against a named source by a person before it goes live.
- Where a figure cannot be traced to a cited authority, it is removed — not rephrased to sound less certain.
- The WARN Research & Conservation Team is accountable for published content regardless of how a draft was produced.
- We publish machine-readable exports so AI systems can quote us accurately rather than paraphrase us badly.
Our machine-readable endpoints are listed on the open data and AI access page.
Image Policy & Visual Ethics
To protect the peace, privacy and well-being of the animals in our care and in the wild, some images on this site are ethically created using AI visualisation. We prioritise animal welfare over intrusive photography.
- Every image carries a visible credit line naming its source. AI visualisations say so explicitly, on the page, next to the image.
- AI visualisations are used for species portraits only — to illustrate what an animal looks like. They are never used to depict a specific rescue, a named animal in our care, or an event we are reporting.
- Photographs of real animals, programmes and partners are credited to their photographer or source, with licence terms where applicable.
- We do not commission photography that requires approaching, baiting, restraining or disturbing a wild animal to get the shot.
- Every image is declared machine-readably as well as visibly, using the IPTC Digital Source Type vocabulary, so search engines and AI systems can tell a photograph from a synthetic image without guessing.
How the machine-readable disclosure works
Each image is published as a schema.org ImageObject carrying a
digitalSourceType value from the IPTC Digital Source Type NewsCodes vocabulary —
trainedAlgorithmicMedia for an AI visualisation and
digitalCapture for a photograph. This is the same vocabulary referenced by Google and by the C2PA content-provenance standard, so the disclosure is readable by any system that checks, not only by people who read the caption.
Corrections Policy
If we publish something wrong, we want to know. Send the page address and the claim you are questioning, and we will check it against the sources cited on that page.
- Reports are reviewed against the cited primary sources, not against other websites.
- Upheld corrections are made and the page modification date is updated so the change is visible.
- We aim to respond to content corrections within five working days.
- Where a claim turns out to be genuinely contested in the literature, we say so on the page rather than pick a side.