September 2019

eBird

Part II: The Checklists

Sept. 14, 2019 update:
The eBird 2.0 Android mobile app now allows offline checklists with GPS location for later uploading.

Even experienced eBird[1] reporters are sometimes unaware of the more subtle aspects of its use, and the options available to them. For beginning users, these details can help them understand the nature of the project and its great importance. In either case, it's clear that "this isn't your mother's checklist"! As we saw in the first article of this series, this is true of how the data can be subjected to analysis, but it's also true of how the sightings are input in the first place.

One threshold question is which birds to include. The simplest and most accurate answer is, "That depends." A parrot in a cage in a pet store? Definitely not. The same species of parrot seen flying free in a flock on Telegraph Hill in SF? Absolutely yes. While answers to this and similar puzzles can be found on the eBird Help pages[2] (a Favorite), how such a sighting is included on a field checklist becomes significant. The reason for this has to do with the dual purpose of the entire eBird effort.

For some birders, eBird is simply an efficient, accessible, online personal checklist archive, with an added opportunity to readily share their sightings. For others, it's a community science project, providing invaluable data to ornithologists, conservation biologists, land use planners, and the like. The birder might only want native species on their life list. The scientists might want all species sighted, whether natives or invasive exotics, to be accurately tallied. How can both be accommodated? At the present, eBird offers no automated way to separate "good" birds from "bad" ones on an individual's life list. What to do?

One simple workaround is to maintain two eBird accounts[3] (Go ahead, it's OK!), reserving one solely for sightings of Chestnut Munias or Eurasian Collared-doves in North America, or any other species deemed to be unwelcome. Some West Coast purists might include Great-tailed Grackles in that group, or even Chestnut-backed Chickadees, if they're fiercely protecting the historic Eastern San Francisco Bay Area aviflora from interlopers. The point, of course, is that the two lists need never be intermingled, and the scientists will still get their desired statistics. If an eBird reviewer does flag the submission(s) as "introduced/exotic", the current protocol is that such birds will not appear on public maps or county lists, presumably to avoid confusing "community" scientists. The invalidated records remain in the dataset, however, and can readily accessed with a Basic Dataset[4] download, which is what the pros do. Like all aspects of eBird, this is a work-in-progess.

A critical component of each checklist is the count! Unlike typical birding field trip lists which merely tally the individual species seen, for eBird science to be useful it makes a difference whether there were over a thousand Sandhill Cranes on a particular Central Valley field on a particular day or just a handful. Which brings up the question of errors and checklist reviews. Your columnist was told that the number of cranes he had reported must be an error that early in the migratory season, even though the birds had been carefully counted one-by-one through a scope. The size of the flock was corroborated within days by several "more reliable" reporters.

Similarly, a flock of Whimbrels seen along one stretch of the California coast had to have been Long-billed Curlews according to another eBird reviewer, although in this case there were photographs documenting the sighting. The camera had been purchased precisely for the purpose of such verification. Most eBirders can share similar stories, and for some the experience is so traumatic that they abandon their participation completely.

What they might not care to admit is the times when they were demonstrably wrong, and when acceptance of their identification and count would have seriously compromised eBird's research. Several dozen Western Northern Waterthrush together in a wetland in Marin County? No, those were female and juvenile Red-winged Blackbirds. A Glossy Ibis at a Wildlife Refuge on Maui? No, it was a White-faced. There's a reason our reports are reviewed! And, according to David Allen Sibley[5], whose story of mistaking a Great Egret for a Logger-headed Shrike is both funny and instructive, we simply can't let ourselves be ego-driven about this aspect of our birding.

It's another issue entirely when checklists are not reviewed or accepted at all, and when significant, accurate, documented sightings do not get added to the searchable database[6] in a timely fashion. This absence of an earlier record can lead some to question their own valid observation at a particular location, particularly if they don't have an unmistakeable image as proof. The backlog which some reviewers/editors have amassed is a current, critical eBird challenge.

Future columns will deal with other intricacies of the enormous web and app-based eBird universe, including "Whose 'Hot Spot' Is It, Anyway?!" What a gift and responsibility we birders have been given.

[Note that the American Birding Association checklist protocols differ from eBird's in a number of ways. ABA online resources will be reviewed in March 2020.]

The Bird Wide Web™ will be publishing a new article on the beginning of each month.

Recent Articles: