Subject: The CBC and useful counts
Date: Feb 22 07:17:15 2001
From: WAYNE WEBER - contopus at home.com


Tweeters (and Vancouver Birders and OBOLinks),

I would agree with Kelly Cassidy only up to a point. Having taken part
in, compiled, and/or edited Christmas Bird Counts for more than 40
years, I feel qualified to speak on the usefulness of CBC data for
tracking population trends.

First of all, the usefulness of Christmas Bird Counts in scientific
research, and in tracking continent-wide population patterns and
trends, has been amply demonstrated, first by Carl Bock and Larry
Lepthein, and since then by dozens of other authors. Certainly, CBCs
are less useful for this purpose than Breeding Bird Surveys, which
were specifically designed to monitor population changes. However,
even BBS data are most useful when data for large numbers of routes
are analyzed together. Data from a single route are too much affected
by changes in observers over the years, and by weather variables, to
give a reliable indication of bird population changes.

It is my observation that the most useful Christmas Bird Counts in
tracking population changes are those with large numbers of observers,
with the same sectors covered by the same people year after year, and
especially with the same compiler for many years. Counts in which the
sectors are assigned differently in different years ("let's see who
shows up and assign parties based on that"), or with frequent changes
in compilers, tend to be far less useful for tracking population
changes.

Despite my earlier comments about Starling counts on the Vancouver, BC
CBC, I think people would be surprised at how well this count actually
reflects population trends in many species over the last 30 years. It
has had 100 or more observers almost every year since 1970, and has
used the same 25 sectors since 1968, many of which have had the same
area leader for many years. (Thanks again to Wayne Campbell for
establishing these 25 areas, and for initially whipping up enough
enthusiasm to swell the number of counters past the 100 mark.)

One of my long-term projects is to write a detailed analysis of the
Vancouver CBC results, using area-by-area results as well as the count
totals. When you separate out the effects of weather and factors like
whether or not Starling roosts were counted. I am certain this count
will be a very useful barometer of population changes for many
species.

Moving back to the continental scene, a few of you may remember when
the late Robert Arbib proposed establishing "ideal model" counts,
which would need at least 50 observers and would record additional
information that would make these counts more useful for tracking
population changes. In exchange for the extra effort, the per-observer
fee on these counts would be reduced. This was an excellent idea, but
after one year, the idea was torpedoed by National Audubon because it
would cost them too much revenue!

The lukewarm support shown for many years by the National Audubon
Society for Christmas Bird Counts has been deplorable. This large,
wealthy organization (of which I have been a member for 25 years)
could easily afford to publish the CBC results without requiring any
observer fee, but many of the directors simply don't understand the
conservation and scientific value of the results. There have been
repeated proposals to stop publishing the CBC results, and even this
year, National Audubon seriously considered making the results
available on the Internet only. Websites come and websites go, but
only a publication will put the results permanently on record. Let's
keep doing CBCs, keep publishing the results, and if possible, bring
back the ":Ideal Model" counts!

Wayne C. Weber
Kamloops and Delta, BC
contopus at home.com


----- Original Message -----
From: Kelly Cassidy <lostriver at completebbs.com>
To: <tweeters at u.washington.edu>
Sent: Thursday, February 22, 2001 6:33 AM
Subject: The CBC and useful counts


> The discussion of the Backyard Bird Count comes at a time when I've
been
> mulling over the Christmas Bird Count (CBC).
>
> Probably everyone who's tried to use the CBC numbers has wished that
someone
> in the last century would have added an unbiased count to the CBC
numbers.
> Because the CBC numbers vary so much with number of birders and
access to
> count areas, it's hard to extract any useful information from them
beyond
> the crudest comparison of species and numbers.
>
> Various formulas have been used to try to compensate for number of
> volunteers, with minimal success. The relationship between
volunteer and
> count is complicated. The difficulty of using CBC numbers is
underscored by
> contrast to the Breeding Bird Survey, which is one of the best sets
of
> long-term ecological data on population trends over a large area.
The BBS
> has a rigid sampling protocol and permanent routes.
>
> I realize the CBC is as much education and public relations as it is
data
> collection. I wouldn't propose eliminating the count as it exists
now or
> changing any of the current CBC rules. I'd suggest instead the
addition of
> an unbiased count on Christmas morning in addition to the regular
count.
> (Christmas morning because traffic would be minimal across the
country
> then.) For example, a 15 mle. road transect running as close to a
diagonal
> as possible across the circle could be permanently established for
each
> circle and sampled by a pair of experienced birders using protocol
similar
> to the BBS (a stop every half mile, counting the number of birds
seen for a
> 3 minute period within a quarter mile, etc.)
>
> The CBC just began its second century. The next count would be a
great time
> to upgrade its methods.
>
> Kelly Cassidy
>