A Gonzaga professor analyzed crowdsourced flycatcher songs to weigh in on an infamous birder debate

click to enlarge JEFF DREW ILLUSTRATION

Jeff Drew illustration

To begin with, Gonzaga environmental studies professor Jon Isacoff offers a caution. He says there may be a point where he has to go off the record to talk about what he really thinks about, say, people who misidentify things based on poor information because they don’t know what they’re doing. He’s the kind of academic who really gets bugged when people get it wrong.

Jon Isacoff

And so even though Isacoff is more a bird fan than a technically trained ornithologist, he’s also North Idaho’s regional reviewer for eBird, a massive crowdsourcing project run by the Cornell Lab of Ornithology. Think of it as roughly like Wikipedia for bird sightings.

“I’m a citizen scientist,” Isacoff says. “That’s what eBird is. It’s a citizen science site.”

The trouble, he says, is some people are less into the science part of it than others. That’s where his sense of precision comes in handy.

“There’s something like 400,000 regular users,” Isacoff says. “If somebody reports, you know, 100 condors in their yard today, it triggers a flag.”

And then it’s his responsibility to wade into the fray, examine their photos, their recordings and their locations, and see if they actually submitted the right bird.

One problem is there’s a whole community of birders who, like Owen Wilson’s character in the bird spotting comedy The Big Year, are incentivized to convince themselves they’ve seen a new bird to add to their list.

“When it becomes 100 percent citizen inputs and zero percent scientists, you’re just getting gobs of bad data,” Isacoff says about eBird.

Play “spot the difference” between a Cordilleran flycatcher and the Pacific-slope flycatcher, and you’ll lose every time.

Back in 1989, the Western flycatcher was split into these two species, but from the outside, there’s absolutely no difference in appearance.

“Even if you were to net the bird and hold it in your hand and measure it with calipers,” Isacoff says, there’s so much overlap in size that you’d rarely be able to identify them digitally that way either.

Instead, there are only two reliable ways to tell them apart, Isacoff says. Get a sample of the bird’s blood, and do a DNA test. Or listen to them sing in the morning.

“This particular genus of birds, unlike other songbirds, their songs are genetically hardwired, they do not learn them from their parents,” Isacoff says. So while the eyes tell you nothing, highly trained ears can listen.

Not just any flycatcher song will do.

“The birds will sing partial songs, bits and pieces of their songs on and off all day,” Isacoff says. Most of them tell you nothing.

But the dawn song — sung around sunrise — comes in three different parts. The second part has the clearest difference. With the Pacific-slope flycatcher on the coasts, there’s a distinct two-note flourish, while with the Cordilleran flycatcher east of the Rocky Mountains there’s “a single slurred whistle,” as bird song expert and author Nathan Pieplow put it.

It’s the difference, as Pipelow writes, between “Teet… Psee-bit!… Chi-peet!” and “Tseet… Pswee-er!… Chippit!”

It’s a subtle difference, the kind detectable by only the top 1 percent of birders, Isacoff says.

In fact, the birding community is absolutely filled with controversy — conspiracy theories! rumors! online fury! — about whether splitting off the species was the right move at all.

“People will say, ‘I’m a splitter, not a lumper,’ or ‘I’m a lumper, not a splitter,'” Isacoff says.

But here’s where all these bird sightings posted on eBird go from being occasional annoyances to genuine assets. Last year, Isacoff turned the 29 uploads of supposed Cordilleran or Pacific-slope flycatcher “dawn songs” in Eastern Washington into a spectrogram — a visual representation of the sound — to detect which samples were of which species.

The answer? For 28 out of 29, they weren’t clearly one or the other. Only one was a pure song, the rest was “an enigma in the middle,” though a little closer to the Pacific-slope song than the Cordilleran song.

“All these people who have ‘life lists’ and ‘year records’ would wander over from Seattle or Tacoma to the Blue Mountains and say, ‘I just recorded a Cordilleran flycatcher.’ They actually didn’t record a Cordilleran flycatcher,” Isacoff says.

They recorded a mixed flycatcher — a mix of the two. So what does that mean for the infamous “lumper vs. splitter” debate?

“There probably are two distinct species genetically, at the extreme east and west of the population,” Isacoff says, but everything in the middle is a hybrid. And Spokane is ground zero for this mix.

And ultimately, that speaks to a far broader scientific debate than simply flycatcher arguments online.

“What we’ve discovered is that the idea of speciation is much more fluid and complicated than what was ever thought before,” Isacoff says. “It’s maybe one data point out of hundreds, but it’s our data point. It’s here.”♦

Article Source: Inlander