The following addresses Bill's experience with querying data in
eBird.

WHAN: "For example, I explored its data for Franklin County, Ohio,
from the year 1900 to the present. I got a list, with abundance
histograms, for 278 species (possible in the county apparently, as
there are no records for most of them)"

Sorry, I do not follow you. What do you mean by "there are no
records for most of them"? A species (or taxa) is listed ONLY if
there is a vetted record within the
 database.

WHAN: and 19 "other taxa," which included stuff like
"warbler sp." and "sparrow sp.", but no taxa like warbler hybrids
or hypotheticals,
.
If there is no entry for a given taxa, none within the birding
community have chosen to enter it. The only warbler hybrids
thus far reported to eBird for the state of Ohio  involve Blue-
winged and Golden-winged Warblers. One of those, a May
08 sighting by Gene Stauffer is shown for Franklin County.

As to the nature of the other taxa, that is the subject of a separate
post.

Hypotheticals? Within a checklist this typically refers to species
for which there are no specimens or photographs,
 merely sight records. That works well within the purview of
records  committees, but how does it work for eBird, the Listserv,
or The Ohio Cardinal.

If I had to venture a guess, I would say ...

Of those observations reported to eBird ... 99.9% are sight records.
Of observations reported to this Listserv ... 99% are sight records.
Of observations printed within journals ... 98% are sight records.

Virtually everything is a hypothetical in that regard.

WHAN: or the reports that were excluded.

A valid point. One always likes to see the whole
data set without  the bias of another editor. I  feel the same way.
No doubt, you could provide an explanation for the basis of
culling observations for  The Ohio Cardinal  while you were
editor. The result would have a been a tiny fraction of  the
data sent to you. No doubt you have the raw files
(photo-copied?  archived? backed up electronically?) and
can readily supply these to  those that request them. For
example, to BSBO so they may identify  additional unpublished
records for the Winter Bird Atlas ... or pinpoint the location to
the necessary quadrangle of those that were printed but where
the specific location was omitted owing to space constraints. I
have spent the better part of the past 12 months reviewing
much of the printed Ohio literature for the Ohio Winter Bird
Atlas, and there is a surprising amount that is  beyond our
reach by virtue of the printed format. Were it pinpointed in
eBird it would have  been a simple matter to make use of the
data.

But I digress. To answer your question as to excluded records,
in eBird,  it took me 3 minutes just now to pull up the 530 Franklin
County records  of which I expressly reviewed and physically
designated as INVALID*.  Four are pending. And yes, I can
supply a word doc. if you wish to sift through exotics,
duplicate submissions, andtranscription errors,  that make up
the bulk of the file.

[*the default condition of all flagged items is INVALID.]

WHAN: "Clicking around gives a local map showing sites at which
included species were reported, when, and by whom. Not all the
locations were accurate."

This is indeed true. And something Ken Ostermiller and I have
been addressing. Given the pace of checklist entry (2200 or so in
the last 7 weeks for example), tackling erroneous placement of data
is quite the challenge. More on our work in a separate post. As I
posted this time last year, I encourage the birding community
to police eBird by bringing these errors to my attention.

BTW, it is a very simple matter for anyone, having identified the
misplacement of their checklist, to relocate it to the correct
locale. Gross misplacement of a checklist, or attaching a broad
location (state and county*) are grounds for invalidation.

*I presently exempt county big days with appropriate effort
information.

WHAN:  A few, very few, of these records go back well before
the  days of eBird.

Very glad you pointed this out. It goes to the crux of my "beating
the drums"for eBird these past 3 years. Historical data ... where
are they? Again, the participation of the birding community, especially
those with the knowledge of where specific records are located,
will be invaluable in fleshing out local and state avifaunas. While
we're waiting, I will continue to enter historical records under the
banner OHIO-ARCHIVIST -VWF ... which includes one of the
earliest  records in eBird ... the only entry for Passenger Pigeon.

On the other hand, Calliope Hummingbird (and dozens of other state
records) will not find a home in eBird without the aid of those with
knowledge of the appropriate geographical placement (nearest
cross-roads works if privacy is an issue) in choosing to enter that
observation. Much of that information IS NOT available in the
printed ornithological record where space constraints force editors
to abbreviate the record. I know I'm forced to consolidate many a
record in North American Birds, as distasteful as I find doing so.

WHAN:  I knew most of the
 contributors, and trusted their
identifications,  but if a lot of beginners start contributing, few
mistaken reports of commoner  species such as those Allen Chartier
describes. They are difficult to vet.

Presumably these same ID problems plague the Christmas Bird
Count yet  this effort seems to have some staying power. I'll touch
on this in a  post on vetting. Worse yet are transcription errors ...
which easily take up most of my time when vetting eBird.

WHAN:  I'm putting together a checklist for the county, for which I
have thus far 329 species, a lot more than eBird,

And will remain the case so long as birders choose not to enter
historical data. It is possible to reconstruct (back to 1900) a checklist,
of all species,  based on frequency. One that with an appropriate
sample size (that participation issue once again), will give a reasonable
hypothesis as to the occurrence of those species across the seasons.

WHAN: and 27 "other taxa," but not
including vague and unsatisfactory categories like "shorebird sp." or
"blackbird sp."

Another excellent point. I keep to a few rules for the
appearance of the checklists generated by eBird (yes
this is my decision not Cornell). I decline to validate exotics.
eBirders are
free to populate their personal lists with Chukars and Lady
Amherst's Pheasant, but I draw the line at the public checklist.
If someone in the future wishes to reverse that policy, it is only a
matter of a few mouse clicks to do so.

For 2 years I adhered to the principle that a taxa
had to be a discrete entity representing at a maximum the genus.
So Greater/Lesser yellowlegs,  peeps sp.,  were acceptable but
shorebird sp. was not. However,  suggestions come in all the
time from end users and despite some misgivings over some of
the choices, I have chosen to respect their submission at least
for a trial period to see whether gull sp., shorebird sp., etc. is
something eBird users wish to see. These extraneous taxa have
only been implemented since spring 2009 and I still consider
them experimental.

WHAN:  It doesn't include extinct and extirpated species. The county
has two records of varied thrush, but they do not appear in these data;
ditto for dozens of other records of unusual species. Why offer a search
that starts at the year 1900 when there are no eBird data for all
species, or for a period no longer than a few years?

Ummmmm...
eBird is a tool by which a database can be created and, to a
limited extent, analyzed. How the Ohio birding community
chooses to wield that tool, and to the extent we take responsibility
for our observations, will dictate the quality and quantity of the
content. One is free to enter any of these
historical records.  I would be pleased were yourself or the say
the Avid Birders took on the task of entering historical published
data for Franklin County. I only ask that any one wishing to do
so contact me first so we don't have overlapping efforts across
the state.

WHAN:
 What's missing from eBird is 200 years of data.
No doubt these can be supplied.

I doubt it.  I just flipped through Wheaton (1882) and found
few specific records. Apart from a few specimen
records, I am aware of exceedingly few published records
prior to 1900 that reference a specific date and location.
Given  its role as a tool for conservation, I am more concerned
about the past 30 years or so of data in eBird.

WHAN: Many of them will come from posts on mailing lists
that later appear in printed publications. I edited the Ohio
Cardinal, the state ornithological journal, for ten years, and
going through such data and those of regional and national
counterpart publications (like N. Am. Birds, etc.), will
presumably always be how bird records are compiled.
they are not lost in cyberspace.

 Are Listserv observations lost in cyberspace? Are CBC data, and
that of a  variety of other databases? How is eBird different in that
respect? I agree, eBird does not serve as a compilation but it can serve
compiler's.

WHAN: And if eBird decides to include this history, this is where
 they will find it.

The listserv .. well yes ... which is why I have appealed to those
taking the time to write up the lists for that venue to consider
saving somebody else (namely me) from having to so. For
example, the same White-winged Crossbill data set that proved
so illuminating was a consequence of my entering 2 dozen
checklists (half of the data) off the listserv. Similarly, with the
Sandhill Crane flight in Dec ... a good many were entered into
eBird, subsequent to my appeal to do so, including
1500 birds not listed within Gabe Leidy's summation. But of
his 6600 drawn from many other sources, the bulk are not
within eBird ... more work for me to track those down ,
determine whether sufficient information is available to
enter the data, and then do it ... all because visualizing such
a flight, and the paths the birds took, is far more informative
than a generic slate of disparate reports.

Now as to the printed ornithological record. Unfortunately no
... this will prove a very limited resource. What % of the 1000's
of records submitted on
report forms finds its way into print? When I was editor of The
Ohio Cardinal, the 55-65 paper reports, plus the listserv could
represent 14-18 thousand observations for a given season. I
would distilled this down to a  spreadsheet of roughly
900 (summer) to 2400 (fall) sightings. From this about half
found their way into print. In other words, 80% of the
ornithological record does not see the light of day.* And
of course, to Dave Slager's point, the effort behind those
observations is rarely made known.

* A wonderful exception are local efforts at summary such as
The Bobolink, The Cleveland Bird Calendar, and Ned Keller's
online database for the Cincy area ... here a much higher
percentage is made available..

WHAN:  EBird has a different and important purpose, the long-term
study of bird populations enabled by large databases of estimated
numbers and locations and seasons of every species.

Exactly so.

WHAN: Finally, I advise against posting one's eBird data
directly to mailing lists in unedited form ... a mailing list is a
poor database, but has other functions.

I agree. And have never held a different position from the time of the
inception of Ohio-Birds.

WHAN:   ... I have no problem that the eBird folks are
playing the drums to get participants, even if it is made to seem
like the greatest thing since sliced bread for almost every kind of
data need.

I think I once said it was the greatest tool
since the Peterson Field Guide for the bird
conservation ... and still stand by that.

eBird allows the common birder to tie their
observations, that presently DO NOT find their
way into any other venue, to specific geographic
locations. For those who tout, rightfully, the
conservations efforts of environmental
organizations through land acquisition,
ask yourselves this ... How does one value that
land from the standpoint of biodiversity? ...
you gather data ... you create lists of fauna
and flora. It is the very foundation of the
conservation of biodiversity. Every county park
system, every land trust, every Audubon Society,
every member of the Ohio Bird Conservation
Initiative, etc. could make use of eBird. Still
unsure ... take a look at the Ohio Breeding Bird
Atlas ... do you participate online? Then you
 are a part of eBird.

So who are the eBird folks. They are several hundred birders,
dozens of whom also participate on the listserv.  They are
Audubon Society groups such as Black River Audubon
diligently canvasing their local Audubon Ohio Important Bird
Areas within Lorain County. They are banding groups, they
are members of the Ohio birding community (and increasingly
Magee bounds tourists) ...  collectively bringing more than
200,000 records annually ...  2200+ checklists at the half way
mark of this winter season.

BTW, that effort ranks Ohio consistently about 13th among the states.

WHAN: One
 other thing. Some people think list postings are just for
reporting bird sightings, but that is far from the case. Look at Rob Thorn's
informative post from last night and imagine how eBird would have told
you that.

Indeed. There is much, much more to the listserv.
Thanks Bill for bringing up these important
points.

cheers

Vic Fazio
Shaker Heights, OH
State Reviewer, eBird, Ohio & Oklahoma
Regional Editor, North American Birds, OH-PA-WV
[log in to unmask]

______________________________________________________________________

Ohio-birds mailing list, a service of the Ohio Ornithological Society.
Our thanks to Miami University for hosting this mailing list.
Additional discussions can be found in our forums, at www.ohiobirds.org/forum/.

You can join or leave the list, or change your options, at:
http://listserv.muohio.edu/scripts/wa.exe?LIST=OHIO-BIRDS
Send questions or comments about the list to: [log in to unmask]