[Taxacom] Read... and believe...

[Richard Pyle]

I already explained the problem in my reply to Francisco, to which you were
replying.

I'll re-state it:

Otherwise, how will the next person know in what sense of the name you
identified the specimen?  Sensu lato? Sensu stricto?  Are you a lumper? A
splitter?  Without pointing to some sort of concept circumscription for a
given name, the next person is left guessing how you perceived that taxon to
be.  If it's a taxon that has been consistently defined in the era in which
you identified it, then no problem.  But many species could be interpreted
as one of several very different concept circumscriptions, depending on who
you are following.

Perhaps an example would be helpful:

The type specimen of Centropyge fisheri is from Hawaii. It is orangish brown
in life, with a white tail.

The type specimen of Centropyge flavicauda is from the South China Sea. It
is blueish brown in life, with a yellowish tail.

The type specimen of Centropyge caudoxanthorus is from Taiwan. It is dark
blueish brown in life, with a yellow tail.

All three species are drab brown with pale yellowish tails in preservative.

Some authors (splitters) regarded the life-color differences as diagnostic,
and warranting recognition of all three at the species level.  Other authors
regarded the life-color differences between C. flavicauda and C.
caudoxanthorus as minor and inconsistent, and regarded the two species as
the same (C. caudoxanthorus being a junior synonym of C. flavicauda). Other
authors regarded the life-color differences between all three species as
minor and inconsistent, and regarded them all as the same (C. caudoxanthorus
and C. flavicauda both being junior synonyms of C. fisheri).

Someone collects a specimen in the Marshall Islands, and identifies it as
"C. fisheri", without any elaboration.

Years later, DNA sequencing reveals a non-trivial and consistent difference
between the populations in Hawaii, the South China Sea, and Taiwan.  The
community consensus converges on the taxonomic opinion that all three
populations should be regarded as distinct species.

Which of the three species lives in the Marshall Islands?  We look at the
specimen, and its drab brown in color.  We try to get a DNA sequence off it,
but the specimen was fixed in formalin.

If the identifier had made it clear that they were following the more
restrictive circumscription of C. fisheri, then we would feel much more
confident that the identification is congruent with the modern
interpretation for that species. But, unfortunately, the person who
identified it just used their own "mental image of a species", and didn't
bother to note what that mental image looked like in the context of possible
synonyms.

If you think this is a rare circumstance; guess again.  This sort of thing
happens a lot; and not just in fishes.

Aloha,
Rich

P.S. Certain aspects of the Centropyge example given have been
fictionalized; to protect the innocent.  


> -----Original Message-----
> From: Stephen Thorpe [mailto:s.thorpe at auckland.ac.nz] 
> Sent: Friday, September 04, 2009 11:18 PM
> To: Richard Pyle
> Cc: TAXACOM at MAILMAN.NHM.KU.EDU
> Subject: RE: [Taxacom] Read... and believe...
> 
> >Exactly.  And that's the problem
> Better clarify for Paul's sake who you are saying has the 
> problem? :) The identifiers or the databases?
> 
> It is interesting to note that I certainly know of cases 
> where if I did identify by checking every character against a 
> description/redescription, I would get the WRONG answer! More 
> often though, if descriptions/redescriptions/keys were all I 
> had to go by, I just wouldn't get any confident answer for 
> many identifications. Direct comparison coupled with enough 
> experience to know what is or isn't important is the only way 
> to go ...
> 
> Stephen
> 
> ________________________________________
> From: Richard Pyle [deepreef at bishopmuseum.org]
> Sent: Saturday, 5 September 2009 9:10 p.m.
> To: Stephen Thorpe; fwelter at gwdg.de; TAXACOM at MAILMAN.NHM.KU.EDU
> Subject: RE: [Taxacom] Read... and believe...
> 
> > This is not how identifications work in practice!
> 
> Exactly.  And that's the problem.
> 
> Aloha,
> Rich=