Kelly and Tweeters,

As I mentioned, determining population trends from BBS data is a
complex process.

Abandonment of routes (or creation of new routes) is potentially
a problem in determining population trends. Other than the Point Grey
route in Vancouver and West Vancouver, which I covered from 1974
to 1994, I am aware of only two other routes in BC which have
been abandoned due to heavy traffic, both in the Greater Vancouver area.

I'm sure that several routes in rural areas have been abandoned as well.
Most roads in BC are logging roads, and tens of thousands of miles
of these have been abandoned over the last 40 years or more. I'm
sure that several BBS routes have become undriveable because of bridges
washing out, roads deteriorating, etc. I know at least one route that was
covered for 2 or 3 years, then abandoned because river noise
interfered so much with observations that useful results could not be
obtained.

BBS coverage in Canada is usually based on one route per degree-block
(a chunk of land measuring one degree latitude by one degree longitude--
about 4000 square miles, depending on the latitude). In more heavily-settled
areas of southern BC, there may be 2 or 3 routes per degree-block.
When a route cannot be covered and must be abandoned, another route is
assigned in the same degree-block. The location of new routes is
determined using random numbers, to try to avoid strong biases toward
or away from certain habitat types.

In the analysis of BBS data, I believe that results are weighted, such that
routes in degree-blocks with 2 or 3 routes "count" for less than those in
blocks with only one route.

However, all these measures attempt only to ensure that habitats covered
are a reasonable sample of roadside habitats in a large area (e.g. a
state or a province.) This is achieved incompletely at best on the BBS,
because roads are always biased in favor of some habitats (farmland, urban
areas) and away from other habitats (forests, wetlands). Selection of
routes has little if any effect on the analysis of year-to-year population
changes. The main effect that dropping or adding a number of
routes might have in an area the size of Washington would be to
affect which bird species had a large enough sample (i.e., number of
routes) to give meaningful estimates of population change. It would
have little if any effect on the direction or magnitude of those changes.

Wayne C. Weber
Delta, BC
contopus at telus.net





----- Original Message -----
From: "Kelly Cassidy" <lostriver at completebbs.com>
To: "'TWEETERS'" <tweeters at u.washington.edu>
Sent: Saturday, July 22, 2006 6:11 PM
Subject: RE: [Tweeters] Breeding Bird Survey population trends



> [mailto:tweeters-bounces at mailman1.u.washington.edu] On Behalf
> Of Wayne C. Weber
> Sent: Saturday, July 22, 2006 4:39 PM

> most years since 1980 (and 2 of them since 1974). (In
> addition, I did one route in Greater Vancouver from 1974 to
> 1994, but have since abandoned it since traffic makes it
> nearly impossible to cover properly now.) The long-term

Wayne's comment about abandoning a route due to traffic brings up something
I've wondered about for a long time. Are routes more likely to be abandoned
if they are in developed areas? Are new routes more likely to be situated
in non-ag and undeveloped areas? Does a non-random abandonment of routes or
placement of new routes skew the results? If routes are abandoned more
often in developed areas and new routes more likely to be established in
less-impacted areas, wouldn't that make the numbers for birds that don't do
well in ag or urban areas look better than they really are as urban areas
expand? Is there any sort of "correction" for these possible biases?

Whatever biases BBS has, however, the methodology is far better than for the
CBC.

Kelly Cassidy
Pullman, where it's hot and dry. The wheat in the surrounding
fields smells like it's roasting.


_______________________________________________
Tweeters mailing list
Tweeters at u.washington.edu
http://mailman1.u.washington.edu/mailman/listinfo/tweeters