On Fri, 30 Sep 1994, Greg Gillson wrote:
> Would the pattern of more rare shorebirds in California have anything to
> do with the latitude that the birds normally fly to?
>
> In other words, are the birds wintering in the right place (latitude), but
> wrong side of the Pacific?
>
> If the birds are wintering in California, but only migrating through
> B.C., Washington, and Oregon, then that would explain the higher numbers
> discovered in California.
>
> --
> Greg Gillson <gregg at tdd.hbo.nec.com>
> Hillsboro, OR
>
That is a plausible explanation, but as Mike Patterson pointed out, you
really have to be able to rule out sampling bias to be sure that the
apparent pattern (more stray Eurasian shorebirds landing in CA than in
PNW) is even real. It seems equally plausible that denser sampling of
shorebird habitats during migration could account for the difference, and
the proximity of big population centers to good shorebird spots in CA
leads to dense sampling. Someone please correct me if I'm wrong, but I
have the impression that Iona in BC reports more rare shorebirds than
single good shorebird localities in WA and OR. If this is true, then it
supports the sampling bias hypothesis better than the overflight hypothesis,
as Iona is close to a big pop. center allowing many birders daily access
(cf. CA). (Unless you want to argue that the birds are stopping in Iona
before moving on, which is getting pretty ad hoc). Even if I am wrong about
Iona, it seems obvious that sampling in CA is on average signicantly
denser than in PNW.

An immediate test of the relative merits of the 2 explanations would be
to plot competent-observer hours per unit of area for localities in PNW
and CA vs. number of rare birds found there. The sampling bias
hypothesis would predict a high positive correlation between the 2
variables, independent of location (e.g., poorly sampled localities in CA
would fall out with localities having a similar level of sampling in PNW).

This is not to argue that the overflight hypothesis is necessarily wrong,
just that it is premature to invoke it as the best explanation without
ruling out sampling bias first.

David Wright
dwright at u.washington.edu