At 09:46 AM 5/19/99 -0700, Kelly Cassidy wrote:
>>They're conservative, and hopefully records accepted by the RBC have
>>a high degree of reliability, but given that only a subset of sightings
>>are ever reported to RBCs, it is clear that they CAN'T be accurate.
>
>
> I don't understand this statement. An individual record may or may not be
>accurate. Whether or not you have all possible records is immaterial to the
>accuracy of an individual record.

Yes. It is material to how closely that set of records matches
the actual distribution of birds on the ground, though.

Take the trivial case, that the subset of records submitted to
the RBC is the empty set. I think we'd all agree in this case
that the set of RBC-approved records would not provide an
accurate snapshot of bird distribution.

Of course, it's not the empty set. I would suggest, though, that
we don't have any metric on what percentage of bird sightings
are submitted, even of rare birds.

My point is primarily that there's really no reason to argue
that the editors of a FIELD GUIDE should base their decisions
on distribution solely on RBC records.

To do so ignores all sorts of other sources of data. Why
would a publisher want to do so?

Unless, of course, they're publishing a list of accepted
birding records.

But that's not what a field guide is.

> All sets of occurrence records are
>subsets of a population, unless you sample the entire population. You're
>saying that RBC records are a subset of a subset, and that makes them
>inaccurate?

More precisely, they're a strict subset of a strict subset...

> One can reasonably argue that a record never submitted to
>anyone, for all practical purposes, never becomes a record.

Again, my statement was made in the context of a claim that
the NGS should rely on RBC records for calculating distribution
data, at least for this one species in this one province. It's
this notion that the publisher must bow to the authority of
RBCs that I disagree with. I have no problem with RBCs
tracking birding records, it's a good thing. I do have a
problem with the claim that's they're the ONLY good thing.

There are other places where records are held, i.e. HMANA
has a national repository of standard observation counts of
migratory hawks. These are submitted records, but not
necessarily submitted to RBCs.

The fact that they're not submitted to RBCs doesn't mean
they're useless for determining the distribution of
various hawk species. They're another source of data
that, say, an NGS editor might want to take into account,
along with MAPS, BBC, BBA, CBC and RBC-approved data.

Maybe Michael doesn't mean to make his claim in such a
strong fashion. He sure seems so, though, at least in
regard to the distribution of Xantu's hummer...



- Don Baccus, Portland OR <dhogaza at pacifier.com>
Nature photos, on-line guides, and other goodies at
http://donb.photo.net