EDIT

ERS (2018-03-28):

Dear all,

to facilitate the discussion about the status "invalid designation", I am continuing in English and put Nick in cc.

The background for this is article 6.3 Melbourne Code: "In this code, unless otherwise indicated, the word "name" means a name that has been validly published [...]".

Consequently, there is no such thing (status) as a nom. inval., because if it is not validly published, it is by definition not a name, but an invalid designation (des. inval.).

In fact, the abbreviation so commonly found in the literature "nom. inval." is not even included in the glossary, nor in the subject index of the Melbourne code (neither is "invalid designation", nor "des. inval.").

In Euro+Med Plantbase, the editors took the code very literally, eventually arriving at the point that such an invalid designation is not even a synonym (which has to be a good name). It is common practice to indicate this, in synonym lists, with a dash before the "name", instead of the "=" which precedes a validly published synonym. However, in most publications, the abbreviation "nom. inval." is commonly used. @Nick, do you have a recommendation how to deal with invalid designations in databases and publications, from a nomenclatural point of view?

In the following example,
http://ww2.bgbm.org/EuroPlusMed/PTaxonDetail.asp?NameId=121643&PTRefFk=7000000
Euro+Med
differentiates between
Basionym; Homotypic synyonyms; heterotypic synonyms; invalid designations and misapplied names.

@Anton, do you know how the status "invalid designation" is assigned in the Berlin Model Portal?

Answer to:

AM: ... Bevor ich es vergesse, möchte ich einen Fall hier aber schon mal erwähnen, und zwar http://ww2.bgbm.org/EuroPlusMed/PTaxonDetail.asp?NameId=7709633&PTRefFk=7100000 , also der erste Fall bei den Synonymen (Bromus ramosus subsp. bikfayensis).
Dieses wird als INVALID DESIGNATION ausgegeben, allerdings habe ich keine Ahnung, woher diese Information stammt. Im BM Editor finde ich sie in keinem Feld. Ist das eine Zusammengesetze Information? Auch wurde nicht der Relationstyp „is invalid designation for“ verwendet. Hat jemand eine Ahnung woher das kommt?
Das wäre wichtig zu wissen, um ggf. ähnliche Fälle auch korrekt behandeln zu können.

We do not have any invalid designation relationships in E+M (BM):

SELECT *
FROM RelPTaxon rel
WHERE   RelQualifierFk IN (8)
ORDER BY RelQualifierFk;

All invalid names (term ids may need to be adapted)

SELECT ns.type_id, ns.citation_id, n.nameCache, n.titleCache, n.id
FROM NomenclaturalStatus ns INNER JOIN TaxonName_NomenclaturalStatus MN ON MN.status_id = ns.id
 INNER JOIN  TaxonName n ON n.id = MN.TaxonName_id
WHERE ns.type_id IN (867, 871, 877, 876, 873, 875) AND nameCache LIKE 'Tara%cup%'
ORDER BY ns.type_id

And the vocabulary (id to be adapted):

SELECT *
FROM DefinedTermBase dtb 
WHERE dtb.titleCache LIKE '%%'
AND dtb.DTYPE = 'NomenclaturalStatusType' AND dtb.vocabulary_id = 19;

Andreas Müller wrote:

All invalid names (term ids may need to be adapted)

I think this is handled sufficiently in #9272.