This is a very promising tool, though some of its limitations could be
problematic to the user and arguably the ebird program...
I ran two polygons, Ft Stevens State Park and the Brownsmead Area. The
50 sites limiter excluded more than 70% of the lists for Fort Stevens.
However, most of the non-hot-spot locations I regularly produce lists
for were among those used, so locally generated constant effort data is
being represented. The potential for exclusion of (arguably more
reliable) constant effort data in some high use locations does exist though.
One of the big minuses is the bar chart data which, as was the case with
Tundra Goose at Nestucca, overly represents rare species. It turns out
(if one believes the chart) that BRISTLE-THIGHED CURLEW is more common
in May than WHIMBREL. And ARCTIC LOONS are more common in the winter
than COMMON LOONS at Brownsmead...
I keep harping on this issue, because these kinds of statistical
misrepresentations create a negative feed-back loop that grows the
errors. I still meet people out at the South Jetty who want to know if
I've seen any Bristle-thighed Curlews today.
Any eBird list for a site gives folks permission to see stuff. So, a
bird like HOUSE WREN, which I have never seen at Ft Stevens, continues
to be reported while PACIFIC WREN is under-reported. The error grows
with every mis-ID, because it's on the list and folks sometimes make
assumptions rather than clean ID's.
Of course, this isn't the fault of the polygon tool, or eBird, for that
matter. They simply reflect the data that's been reported to the
system. We all need to be aware that the data is often kind of squishy
and those folks who want to use this (or any communally generated
data-set) as a tool for doing science or making management decisions
need to run it by some knowledgeable experts to fine-tune the
message these list represent...
--
Mike Patterson
Astoria, OR
The phenological margins
http://www.surfbirds.com/community-blogs/northcoastdiaries/?p=2753
OBOL archives: www.freelists.org/archive/obol
Manage your account or unsubscribe: //www.freelists.org/list/obol
Contact moderators: obol-moderators@xxxxxxxxxxxxx
