Data provenance
Sources underpin our profiles, articles and guides, but their value depends on what they document, how the data were collected and where the findings apply. This page explains our criteria; individual references belong with the content they support, not in a general directory.
A source may document a scientific name, an observation, an ecological process or a study finding. These are different kinds of evidence. We distinguish original research, datasets, institutional records and summaries. Finding or linking a page does not verify every statement on it: we examine the relevant passage and record the claim it can actually support.
We favour primary research, versioned taxonomic catalogues, public institutions and datasets with accessible methods and metadata. Reviews help put findings in context; popular accounts can guide research but cannot be the sole support for a sensitive or precise claim. We consider the publisher, date, method, geographic scope and independent confirmation, avoiding several copies of the same underlying mistake.
Species profiles check the scientific name, rank and accepted status in the consulted version of a taxonomic resource. Synonyms and classification changes are documented without treating two names as two species. Common names depend on language and region; a literal translation does not establish local usage. Unresolved disagreements about identity or classification remain explicit rather than being presented as certainty.
An occurrence record documents an observation at a place and time, not an entire range. Missing records may reflect limited sampling. Habitat descriptions distinguish general conditions from regional examples. Conservation categories need their issuing authority, geographic scope and assessment date: a global assessment does not describe every local population. We do not extrapolate a study to another place, season or species without evidence.
Wikipedia editions in several languages help locate names, concepts and bibliography and reveal differences in terminology. Important claims are checked against the cited underlying material or independent documentation. A translated Wikipedia passage alone cannot establish a regional name, distribution or biological fact. If its references do not support a statement, we seek stronger evidence or omit the claim.
We check particular claims, not only a website’s reputation. The evidence must address the organism, location, period and phenomenon described, and its method must support the conclusion. When several sources are required, they should represent genuinely independent evidence rather than repetitions of one report. Conflicting findings call for an explanation of limits or further review, not the most striking interpretation.
Accepted names, conservation assessments and observation databases change. Where available, we retain dataset version, record identifier, responsible organisation and access or assessment date. A FaunaGuide review date says when a page was checked, not that every fact will remain current indefinitely. If a source disappears, changes or loses context, we reassess the claim connected to it.
Biological evidence and reuse rights are separate questions. Public or open data do not automatically license accompanying prose, images, video or sound. For each asset we retain the creator, original page, licence, attribution conditions and identification context where applicable. Conceptual artwork is labelled as such and is not evidence of a species’ appearance. An ordinary article may be published without an image when its format allows it.
We explain limits of identification, geographic scope or data quality when they affect interpretation. A photograph may illustrate a species without proving it occurs at the pictured location. Material uncertainty blocks conclusions that depend on it. We review content when classification changes, new evidence emerges or a documented correction arrives; new information is checked before a page is updated.
This page describes general practice rather than listing links. Relevant sources appear with the species profile, article or habitat they inform, so a reader can examine the evidence next to the claim. This avoids mistaking a general catalogue for proof of every statement. Readers can request a documented correction if attribution, a link or the scope of a conclusion is wrong.