Further to this item. I am using a query which returns all occurences at a given location or its sub-sites (location keys "DSS006280000000V", "DSS006280000000W", "DSS006280000000X") of all taxa with statuses which appear in my custom made taxon designation set (key 'DSS0062800000001').

The query looks like this:
SELECT DISTINCT INDEX_TAXON_NAME.RECOMMENDED_TAXON_LIST_ITEM_KEY AS TLIK, INDEX_TAXON_NAME.PREFERRED_NAME AS [Scientific name], INDEX_TAXON_NAME.COMMON_NAME AS [Common name], INDEX_TAXON_NAME.SORT_ORDER
FROM ((INDEX_TAXON_NAME AS INDEX_TAXON_NAME_1 INNER JOIN (((TAXON_OCCURRENCE INNER JOIN SAMPLE ON TAXON_OCCURRENCE.SAMPLE_KEY = SAMPLE.SAMPLE_KEY) INNER JOIN TAXON_DETERMINATION ON TAXON_OCCURRENCE.TAXON_OCCURRENCE_KEY = TAXON_DETERMINATION.TAXON_OCCURRENCE_KEY) INNER JOIN INDEX_TAXON_NAME ON TAXON_DETERMINATION.TAXON_LIST_ITEM_KEY = INDEX_TAXON_NAME.TAXON_LIST_ITEM_KEY) ON INDEX_TAXON_NAME_1.RECOMMENDED_TAXON_LIST_ITEM_KEY = INDEX_TAXON_NAME.RECOMMENDED_TAXON_LIST_ITEM_KEY) INNER JOIN TAXON_DESIGNATION ON INDEX_TAXON_NAME_1.TAXON_LIST_ITEM_KEY = TAXON_DESIGNATION.TAXON_LIST_ITEM_KEY) INNER JOIN Taxon_Designation_Set_Item ON TAXON_DESIGNATION.TAXON_DESIGNATION_TYPE_KEY = Taxon_Designation_Set_Item.Taxon_Designation_Type_Key
WHERE (((SAMPLE.LOCATION_KEY) In ("DSS006280000000V", "DSS006280000000W", "DSS006280000000X")) AND ((TAXON_DETERMINATION.PREFERRED)=Yes) AND ((Taxon_Designation_Set_Item.Taxon_Designation_Set_Key)='DSS0062800000001'))
ORDER BY INDEX_TAXON_NAME.SORT_ORDER;

This query consistently returns Centaurium erythraea, even though RedList_GB_Pre94:EX (the only status attched to this taxon as far as I can tell) does not appear in my custom designation set.

What is wrong with my query?

And yes I have rebuilt the indeces.

A year later and it's still extinct?

Bumping this one to be sure you have seen it Chris. Nearly three years after my original post it is still apparently extinct. Apart of course from the five times I have recorded it this year (and the several thousand records on the gateway).

Funnily enough, the Gateway also lists it as pre 1994 extinct, despite holding all those post 1994 records of it.

I would be willing to bet that it is at least partly the fault of the Rec3.3 list again.
Looking in the taxon dictionary the rec 3 list has three child taxa of C. erythraea. Two are hybrids (which as we have discussed before, are not really children in the taxonomic sense and the third is a variety var capitatum. Interestingly this variety has C. erythraea listed as one of its synonyms. Not quite clear how a taxon can be a distinct entity at the same time as being synonymous, unless there is only one variety, but there you have it.

However that doesn't alter the fact that it appears in the JNCC list as extinct.

Not sure I follow you Chris.

We have to use statuses and we accept the difficulties inherent in their change. The problem is with the synonymy surely.

If there was considered to be (pre 84) an entity called C. latifolia which was considered to be extinct, but it was later decided that this entity was not in fact a species, but fell within C. erythraea, then it should have been demoted to a subspecies or more likely a variety, not made synonymous.

The problem I have is that, even though we explicitly exclude the pre-84 red list statuses from our reports and queries, we cannot apparently rid ourselves of this non-notable taxon, except by post processing the output from recorder. This is because all our queries have to be designed to pick up all the synonyms.

In fact the solution we have adopted is to have a local-non-notable status which we apply to all such taxa which report anomalously. All our queries therefore need to exclude all taxa with this status. This does however represent just one more thing we have to bear in mind when writing already complex queries.

Or that's how it seems to me anyway.

In looking at the JNCC list in the Taxon_designation table outside Recorder you do need to take into Account the date_to field. If this is not null then the designation doesn't  apply. Up to Dictionary 0000000Y it looks like the NameServer links were causing table Index_Taxon_Designation to pick up  Centaurium erythraea from the entry for  Centaurium latifolium,  but as far as I can see this is fixed in Dictionary 00000010 - does anyone know different ? .