Note: to see Part 1, click here.

Sherlock Holmes famously lectured Watson, “How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?” Science might strive for such certainty, but it’s a lofty goal and we usually have to settle for second best. So instead, “when you have eliminated the impossible and rejected the least likely, whatever remains best justified, must be regarded as probably true.” Not a phrase elegantly turned, but a reasonable sentiment . . . and one that applies to our December empid.

I originally noted that had I caught the questionable bird in August, I would have banded it as a Yellow-bellied Flycatcher without question. Why? Because Yellow-bellied are common in our area at that time, and the other look-alikes would have been wildly improbable. Catching the bird in December altered this calculus, as in this month Yellow-bellied Flycatchers are even more wildly improbable than the Pacific-Slope member of the Western complex. The photos below show the number of accepted eBird reports of both Yellow-bellied Flycatcher and Pacfic-slope Flycatcher for the month of December (in all years!).

Yellow-bellied Flycatcher has a paltry three reports on what we can charitably call the northeastern quarter of the continent: a single bird each on Long Island in 2009, Ontario in 2015, and Ohio in 2017.

By contrast, Pacific-slope Flycatcher has 8-12 reports: Lancaster PA in 1990; Lancaster PA in 2001; Harrisburg PA in 2012; a bird each in Connecticut and Reading PA in 2015 (the PA bird showed up 8 hours before the last sighting of the CT bird); two or three individuals in Massachussets, Philadelphia PA and Cape May NJ in 2019 (it is clear the Philly bird is different from the MA and the NJ bird(s), but it’s possible the MA and NJ sightings are of the same bird; and Massachussets in 2020. There currently also appear to be Pac-Slopes in Maryland, Nova Scotia, and North Carolina.

So, the probability seemed to be in our favor. Enter perception bias. It’s one of the hardest things to avoid in birding – letting your expectations color your perceptions. There are any number of minute differences between the Yellow-bellied Flycatcher and the Pacific-slope Flycatcher, but they are small and subject to interpretation. Both have almond-shaped eyerings, but the Yellow-bellied is said to be more rounded at the back of the eye, and the Pac-slope more pointed, for example. So how does our bird look? Judge for yourself.

People commenting on facebook and in messages were divided:

  • “The white ring surrounding the eye on a western flycatcher would come to a tear-drop point as on this bird, so this guy looks to be good for a western flycatcher.”
  • “It looks suspiciously YBFL . . . Eyering looks round, narrow and yellowish.”
  • “It looks perfectly round in the left eye, and tear dropped in the right…”

In fact, people were pretty split on the bird’s appearance overall!

  • “Looks good for “Western”. Here in _____ I’d have just marked this down as a Pac Slope.”
  • “It’s not quite yellow enough for a Yellow-bellied.”
  • “Pacific Slope is one of our most common birds at _____, and if I had this bird, I’d be going “hmmmmm, this looks different!””
  • “It looks like a longer primary projection than I’m used to on Western Flycatchers.”
  • “I’d not be quick to rule out Yellow-bellied.”

An informal tally seemed to suggest that about 75% of the East-coast commenters thought it was a Yellow-bellied Flycatcher, whereas West-coast commenters were evenly split.

Characteristics like “yellowishness” and “rounded” are notoriously subjective, and even experienced birders can be misled by their expectations. Measurements might be a more objective guide, and they are . . . when you take them without detailed knowledge of the expected ranges. For example, suppose you know that Yellow-bellied Flycatchers have an expected bill length of 7.0-9.4mm, and Pacific-slope Flycatchers have an expected length of 7.7-9.2mm. And suppose you measure the bill length at 7.4mm. If you expect or hope the bird is a Pac-Slope, it’s easy to convince yourself that you measured wrong. So you measure again. And maybe again and again, until you get the measurement that is “obviously” right. In cases like this, it’s important to take extra pains to take the measurements in as neutral a way as possible, ideally without having the expected ranges in front of you and without looking at the dial on the calipers until the measurement is complete.

So – where are we? Timing is one bit of evidence, and it favors Pacific-slope. Appearance is a second bit, and most folks seem to think it favors Yellow-bellied. Measurements are a third bit, and they were no help at all. DNA will be the most conclusive bit of evidence, but we may have to wait weeks or months for the analysis. Is there nothing else we can consider now? Ah . . . but there is.

Recall that I was able to take a video of the bird vocalizing. A friend of the station, Nick Kachala, took the the sound file and created spectrograms that give us a sort of picture of what the calls “look” like. Here are his spectrograms . . . compare them with these spectrograms of Yellow-bellied Flycatchers, these spectrograms of Pacific-slope Flycatchers, and these spectrograms of Cordilleran Flycatchers. Which do they match the best? (Note that there are lots of different calls and sounds represented in those recordings – you’ll need to scroll through several to find a picture that “matches.”)

With this evidence in hand, and with the physical appearance favoring Yellow-bellied over Pacific-slope, we believe that this is probably a really (really really!) late Yellow-bellied Flycatcher. In some ways, that’s disappointing, but in some ways it’s also more interesting.

We’ll still send the DNA out to be analyzed (because who knows . . . maybe this is actually a Pac-slope / Yellow-bellied hybrid!), but for now, our best evidence suggests it’s a plain-old Yellow-bellied who is taking migration at a more leisurely pace.

-Andrea Patterson