[Taxacom] How to refer to unplaced taxa

[Doug Yanega]

I concur with Stephen; most databases (at least those that try to adhere 
to Darwin Core standards) have a hierarchy of taxonomic fields, and you 
simply enter names accurate to whatever level you are certain of. Some 
specimens may only be IDed to order, some to family, some to subfamily, 
etc. - and a well-designed database should be able to accommodate. The 
rule of thumb is one field for every rank in the hierarchy *or if not* 
then make sure that a given combined field can be easily parsed out 
should the need arise. The database I use has only three combined fields 
out of the eight in the hierarchy: (1) order (2) any ranks between order 
and superfamily [typically just suborder] (3) superfamily (4) family (5) 
any ranks between family and genus [typically just subfamily] (6) genus 
(7) subgenus (8) species plus subspecies (includes author name) 
[typically just species]. The "species" field can be a provisional name, 
like "n .sp." or "cfr. punctatus", as long as it is unique and 
consistent within your system. Note that the hierarchy should have at 
least a few nested "auto-fill" features, so (e.g.) if you enter a genus 
name, it will fill in the entire hierarchy above that rank. In the list 
above, genus and family are the ranks which cause an auto-fill cascade 
in our database; the database therefore contains three linked tiers of 
taxonomic files - one that goes from subspecies to genus, another from 
genus to family, and another from family to order. It can be designed 
differently, of course, but three tiers seemed like an efficient compromise.