Corrections Policy

A reference site compiled from published sources will contain errors. Assessments are revised, figures are transcribed incorrectly, a range is read from the wrong column, and a sentence written accurately in one year becomes wrong in the next. The useful question is not whether mistakes occur but what happens when one is found.

This policy sets out how to report an error, how reports are handled, the difference between a correction and a clarification, and how confirmed changes are recorded. Reader reports are treated as part of normal editorial work. They are the fastest way most errors on a site like this one get caught.

How to report an error

Error reports go to [email protected]. A report is easiest to act on when it identifies the page, quotes the specific statement or figure in question, and explains what appears to be wrong. If a published source supports the correction, naming that source shortens the process considerably, because verification can start from the same material rather than from a search.

Reports do not need to be formal, and there is no expectation that a reader will have a citation ready. A note saying that a depth figure looks implausible is a perfectly good report. It gets checked against the source either way, and being unable to name a source is not a reason to stay quiet about something that looks wrong.

What happens next

Every report is read. The first step is to reproduce the problem against the source cited on the page: the figure is checked in the original assessment, agency account or reference database, and the conversion is recalculated. If the source supports the current text, the report is closed and the reader receives an explanation of where the figure came from, which is often more useful than a change would have been.

If the source does not support the text, or if the source itself has been superseded, the page is corrected. Where the same figure appears on other pages, in a comparison view or in a guide, those are corrected in the same pass, since a fix applied in one place and not another is worse than no fix at all.

Corrections and clarifications

A correction addresses a factual error: a wrong number, a misstated conservation category, an incorrect scientific name, a bad conversion, a misattributed behavior or a source that does not say what the page claims it says. Corrections change the substance of what was published, and they are recorded as changes to the record.

A clarification addresses text that was accurate but easy to misread. Adding the note that a maximum speed is a brief burst rather than a cruising speed, or specifying that a length range describes mature females, does not correct a false statement. It removes an ambiguity. Clarifications are noted as clarifications so that a reader can tell whether the earlier version was wrong or merely unclear.

How changes are recorded

Confirmed corrections carry a date. The profile's review date is updated, and where a correction changes a figure a reader may have relied on, the page notes what was changed and when. Substantive corrections are described in enough detail that someone who read the earlier version can see exactly which statement moved and in which direction.

Routine editorial work is not logged this way. Fixing a typographical error, rewording a sentence for readability or adjusting a heading does not alter what the page asserts, and recording every such edit would bury the corrections that matter among changes nobody needs to know about. The test applied is simple: if the change alters a claim a reader might have relied on, it is recorded, and if it does not, it is not.

Errors are fixed, not erased

The commitment here is straightforward: when this site gets something wrong, the error is corrected openly rather than quietly deleted. Removing a page or silently swapping a number leaves a reader who acted on the earlier version with no way of knowing that anything changed, and it hides the site's error rate from the people best placed to judge it.

This applies to uncomfortable corrections as much as to trivial ones. If a figure was overstated, the record says it was overstated. If a claim rested on a source that turned out not to support it, the record says that too. A visible correction is evidence that the process works, and treating it as an embarrassment to be hidden would defeat the entire point of citing sources.

Disputed figures that are not errors

Some reports concern figures that are contested in the literature rather than wrong on this site. Maximum sizes for deep-water squid, lifespans for slow-growing sharks and burst speeds for large fish all vary between reputable sources because the measurement methods differ. In these cases the page is not corrected to a single preferred number.

Instead, the disagreement is documented. The note attached to the measurement records the competing figures, explains why they differ, and identifies which one the profile treats as best supported. That is a more honest outcome than choosing a side, and it leaves a reader able to follow the citations and judge the question independently.

Timing and replies

Reports are handled in the order they arrive, with clear factual errors on published figures taking priority over questions of wording or emphasis. Verification depends on how quickly the underlying source can be checked, which is fast for a conservation assessment and slower for a claim that traces back through several publications.

Every report that includes a return address receives a reply, whether or not it results in a change to a page. A report that turns out to be mistaken is answered with an explanation of where the published figure came from, which is often the more useful outcome for the reader who raised it. Correspondence goes to [email protected], and the contact page describes the other routes available for source questions, permissions and general inquiries.