Date: 7/5/17 6:16 pm From: Joel Geier <joel.geier...> Subject: [obol] Re: Reporting of "slash" category Townsend's Warblers in the Oregon
While I share Dave Irons' opinion that Townsend's Warblers are scarce or perhaps even nonexistent in the Coast Range after May, I differ with Dave in my opinion as to where that opinion should be enforced, in lasting datasets.
Reporting a bird as Whatsit/Whatsnotit on a checklist does not preclude filtering such observations based on the opinions of knowledgeable observers like Dave, when it comes to scientific analysis. It's simply a matter of inserting a bit of code that says, in effect:
"if reported as (x/y) where y is overwhelmingly expected, then assume y."
This is not hard to do. It also has the advantage of making these assumptions (even when well-founded, as I consider to be the case in this instance) explicit in the analysis, rather than implicit as a function of what amounts to browbeating of observers.
eBird utilizes the observations of amateur observers with widely ranging levels of experience. I just got done writing up the summer edition field notes for Corvallis Audubon, where I chose to ignore a report of a regional rarity (Cassin's Finch) with no meaningful details by an observer who acknowledged in his profile that he was only up to 50 on his life list, yet that report seems to have slipped past the county editor.
Observer uncertainty is part of the terrain, and eBird editors had better figure out a way to live with it. Let observers who are unsure express their uncertainty. Correct it by algorithms in your analysis. Making upstream "corrections" by influencing observers to report one way or another, or to suppress their doubts, leads to statistical bias that is far more difficult to correct by algorithms.
Good birding, Joel
On Wed, 2017-07-05 at 01:07 -0400, <obol...> wrote:
> From: David Irons <LLSDIRONS...> > Subject: [obol] Reporting of "slash" category Townsend's Warblers in > the Oregon > Date: Wed, 5 Jul 2017 02:22:50 +0000 > > Greetings all, > > This is a bit of an esoteric topic. I am in the process of doing some > eBird reviewing for Washington County and I have come across several > recent (JuneJuly) reports of warblers reported list birds as either > Townsend's/Hermit or Townsend's/Black-throated Gray. These reports > involve heard only (singing) birds. The songs of this threesome of > species are highly variable geographically in terms of tone and > pattern. Local dialects are many, particularly for Hermit and > Black-throated Gray. In many cases it is hard to put a name to unseen > birds. This is where is it is important to understand range and > probability. > > > Hermit and Black-throated Gray Warblers are abundant breeders in the > Oregon Coast Range. Hermit Warblers predominate at higher elevations > and upslope habitats that are conifer dominant. As a general rule, > Black-throated Grays tend to be found at lower elevations in habitats > with a strong mix of red alder and big leaf maple interspersed with > conifers. This includes streamside sites, where red alder is often the > dominant tree species. > > > We are near the southern end of the breeding range for Townsend's > Warbler, which does not regularly summer or breed in Oregon's Coast > Range. Some migrants move north through the Coast Range, but after > about the third week of May I would be surprised to find a Townsend's > anywhere in Washington County. A JuneJuly bird would be most > surprising in the Coast Range, or in the upslope areas closer to > Portland (Forest Park and the West Hills). > > > In my opinion, using either the Townsend's/Hermit or > Townsend's/Black-throated Gray slash categories for the reporting of > unidentified singing Setophaga warblers in these habitats is > potentially misleading, as it suggests the possibility or likelihood > that Townsend's is present in the Coast Range and surrounding ridge > lines at this season. In my experience, they are not. For heard only > Coast Range warblers whose songs can't be identified, I would > recommend leaving them either unreported or assigning them to the > Hermit/Black-throated Gray slash category. Use of any slash category > that includes Townsend's is the least accurate way to report such > birds. I can't in good faith validate such sightings, unless the > observer supplies an audio recording that clearly suggests Townsend's > Warbler. > > > Some sources suggest that Townsend's and Hermit songs are hard to > distinguish and this may well be true in the contact (hybrid) zones > where both species breed. There is a significant zone of breeding > contact, which stretches along the crest of the Cascades (mostly > higher elevation eastern slopes) from Deshutes County, Oregon north to > Yakima County, Washington. Hybrids are fairly easy to find in this > contact zone and I would expect that sorting out the vocalizations of > unseen birds in this overlap zone is more of a challenge than it is > elsewhere. Away from this hybrid zone and with seen birds that are > apparent non-hybrids, I have never had much trouble telling Townsend's > and Hermit songs apart. To my ear, Townsend's songs sound a bit > slower, more slurry and generally less strident. Their songs generally > lack the two hard notes at the end that I hear in most iterations of > the Hermit Warbler song. Hermits tend to sing faster in crisper, > slightly higher-pitched notes. They rarely sound " > drunk," which is how I often characterize the pattern of Townsend's > Warblers, particularly those heard singing away from breeding grounds > and during migration. > > > I post this in hopes of nipping this reporting trend in the bud and > saving us (Shawneen and me) some work. Shawneen and I like to query > observers and get their feedback before invalidating sightings out of > hand. > > > Dave Irons > > Beaverton, OR > > Washington County eBird Reviewer