On 5 Apr 2004, Nancy wrote:

> I have a question about protocol. If you hear a bird and positively
> identify the call but do not actually see the bird, does it count as a
> bird count?

If it counts as an identification, I guess that depends on whose
rules you're following. If it counts as a number in a census or
tally, well, that depends too, but more on your confidence and how
you're moving about. For example, I have a number of regular routes I
walk and I do count "heard only" birds. This is especially easy on a
more-or-less straight line transect. On circuitous routes I have to
remember in what general area I heard a species so I don't double
count it. You get better with practice -- 30 years ago I would not
have dreamed of attempting this. When I'm really confused by
flocking, unseen, twittering birds I just record the number of, say,
Golden-crowned Kinglet flocks I heard. This is not an accurate count
of a species, of course, only of the individuals singing or calling.
Birds that move around a lot, like American Crow, can create their
own problems, especially if they're calling infrequently.

> I am puzzled as to how anyone can count a group of 100 birds or more
> if they are moving? What are the techniques for counting large groups
> of birds? Nancy Renton, WA.

Counting large numbers of birds in flight _is_ really tricky, but it
can be done. Start first by getting comfortable with counting large
flocks on the ground by 10s or 25s or 100s or whatever. It's not an
exact number, but there are ways to increase your precision. The most
obvious way to to repeatedly count the flock and take an average. A
better way is to vary how you count: from left to right, then right
to left, etc. This technique helps overcome any bias to lump the same
portions of the flock as another 25 or 100 or whatever. Last Sunday,
for example, I was confronted with having to count a large, distant
flock of waterfowl on Nisqually Delta. I first counted them from left
to right and got 470, then from right to left I got 440; I used 450
as the 'count.' I then scanned the whole flock quickly and decided it
was 95% Green-winged Teal and 5% American Wigeon, Mallard, and
Northern Pintail. Rounding off, I settled on 430 GWTE, and 6 each of
the other three species. How accurate are those numbers? Probably
within 10 percent for the GWTE and 20% for the others.

Once you get comfortable counting stationary or slow moving birds
this way, it's less daunting to try the same with birds in flight.
Start with flocks that are moving in a steady direction: get in front
of them with your binocs or scope, hold steady, and count the birds
as they stream past. Then, when you really feel self-abusive you can
move on to erratically swirling flocks of shorebirds! ;-)

In a technical sense, we're talking about precision (consistency,
repeatability,) accuracy (correctness, closeness to the true value),
and uncertainty of measurements. You can improve your precision by
repetition, but still be inaccurate if you haven't mastered the
technique of counting by groups. If the density of birds throughout a
flock is more-or-less uniform, then you can improve your accuracy by
counting to individual birds in one "binoc-view," and then counting
how many "binoc-circles" cover the flock.

This is not especially accurate, but it's better than a guess, and
repetitions will give you a hint as to your precision. Play with it
long enough and you'll soon discover your own "tricks."


*****************************
Douglas Canning
Olympia, Washington
dcanning at nisquallyestuary.org
*****************************