GeoCoding et cetera

[christian thompson]

The real problem is that many "bioinformatics" people think retrospective
geocoding is something simple that you can automate or turn over to
non-specialists to do.

The bottomline is like for identification, the geocoding of old label data
can only be done by specialists who know the taxon and the history of those
who have work on and collected the specimens.

The examples of both Robin Leech and Robin Panza illustrate this.  And most
of us can come up with many more examples, but instead of investing in
supporting the existing taxonomic specialists and training a new generation
of them, most bioinformatics projects seem to  seek "mine" information
without them and are destined to get garbage as that is what they enter into
their databases.

The Entomological Collections Network years ago stated that retrospective
data capture from specimens should only be done as part of the revisionary
work of a specialist, when the identification can be verified and specialist
can provide the best point estimation (or rectangle) of the original
collection.

There is no cheap, easy or quick way to get quality biodiversity
information. And as the ad go, you can only get it the old fashion way, by
hard work!

Also, let me clearly state, we fully endorse PROSPECTIVE specimen data
capture.  All new collected material should have coordinates and a precision
factor if necessary.  And also specimens need to be  individually barcoded,
etc.


F. Christian Thompson
Systematic Entomology Lab., ARS, USDA
Smithsonian Institution
Washington, D. C. 20560-0169
(202) 382-1800 voice
(202) 786-9422 FAX
cthompso at sel.barc.usda.gov
visit our Diptera site at www.diptera.org

>>> Robin Leech <releech at TELUSPLANET.NET> 02/13 10:16 AM >>>
There is another problem not touched on so far, and that is that late in
the
1800s and early 1900s, collectors were all over a particular state, say
California, and many other western states in the western U.S.  The
collectors sent their material to a person in Los Angeles.  When the
material went to a museum, the locality data read "Los Angeles, Cal.".
The
specimens could have come from anywhere in the western U.S.
Robin Leech
----- Original Message -----
From: "Panza, Robin" <PanzaR at CARNEGIEMUSEUMS.ORG>
To: <TAXACOM at USOBI.ORG>
Sent: Tuesday, February 13, 2001 7:55 AM
Subject: Re: geographical coordinates and accuracy


> However, while 50m may be significant to plants and sessile animals, it
is
> utterly meaningless to birds, so we'd have to have quite a range of
distance
> codes to be able to share them among taxa.  The kind of accuracy problem
I
> deal with is the specimen labelled "British East Africa", or
"Pomerania".
> Or "Pittsburgh", which may mean within the city limits or somewhere in
the
> general vicinity (there's a lovely term!) of the city.  Or "Walters,
CA",
> which I finally tracked down to a RR stop that is now underneath the
Salton
> Sea (California, USA)--obviously, that species of bird is no longer to
be
> found within 50m of *that* locality.  Or "Palestine", which may become a
> valid name again, but with different borders than it had at the time
this
> specimen was collected.  I care about the specimens all labelled the
same
> because that's where the collector had his base camp, although he
travelled
> as much as 25 miles in every direction from there to do his collecting,
so
> we have several subspecies of the same species, all labelled with the
same
> locality.  You're worried about accuracy to within 10m?
>
> With GPS units becoming so affordable, it's getting easy to forget that
> those of us with collections >100 years old have to have a database
system
> that allows for both modern and ancient accuracy levels.  I *don't know*
the
> accuracy of locality labelling for the vast majority of our nearly
200,000
> specimens, but I *do* know that it varied wildly.
>
> Robin
>
> Robin K Panza                         panzar at carnegiemuseums.org
> Collection Manager, Section of Birds          ph:  412-622-3255
> Carnegie Museum of Natural History       fax: 412-622-8837
> 4400 Forbes Ave.
> Pittsburgh  PA  15213-4008  USA
>
> -----Original Message-----
> From: Steve Shattuck [mailto:steves at ENTO.CSIRO.AU]
>
>  Arguably a standard set of codes could be
> developed (1 = within 10m, 2 = 11-50m, etc) but given the independent
nature
> of taxonomists it's unlikely this would ever be agreed on never mind
> accepted and implemented.
>