Browsed by
Category: Identification

Black and Pigeon Guillemots

Black and Pigeon Guillemots

Black Guillemot in "hunch-whistle" posture, Moville, Ireland, 5/22/2007. Photo by Sean Mack (CC 2.0).
Black Guillemot in “hunch-whistle” posture, Moville, Ireland, 5/22/2007. Photo by Sean Mack (CC 2.0).

“No differences known,” says the Sibley guide about the voices of Black and Pigeon Guillemots.

This isn’t quite true.  But the differences certainly are not widely known. Since these two species are impossible to mistake for other alcids either visually or vocally, and because they barely overlap in range, most birders have neither the opportunity nor the incentive to listen for vocal differences between them.  But the differences are quite striking when you know what to listen for.

Each guillemot species has two types of long, complex vocalization, along with some shorter calls. The long, complex vocalizations are rather plastic and grade into one another within each species.


It’s an open question whether the term “song” is appropriate for any guillemot vocalizations, but for now, I’m using it to describe the most complex and most stereotyped sound in each species’ repertoire, often given from the nest.  In both Black and Pigeon Guillemots the main part of the “Song” is usually preceded by a high, rapid series of chip-like notes.  Black Guillemot follows up with a series of couplets that can recall the “squeaky wheel” song of Black-and-white Warbler.  Sometimes this is followed by chip-like notes again. In Pigeon Guillemot, by contrast, the initial chips culminate in a single longer whistle and then a fast, slightly falling trill:

Black Guillemot "song"
Black Guillemot Song, Meat Cove, Nova Scotia, 6/27/2007 (LNS 133325; click image to listen. Spectrogram is from the tenth minute)

Pigeon Guillemot "song"
Pigeon Guillemot Song, Bandon, Oregon, 6/24/2000 (LNS 109262; click image to listen).

Overall, Pigeon Guillemot’s song is much faster. Black Guillemot rarely, if ever, gives a series of notes fast enough to be called a trill.

Seet Series (“Hunch-whistles”)

In addition to what I’ve called song, both species of guillemots give another longer but simpler vocalization that may have a more aggressive function, and which is usually given in a “hunched” posture like that shown in the photo above.  (“Songs” can be given in a somewhat similar posture.)

In Black Guillemot, the Seet Series or “hunch-whistle” is a series of very high couplets again reminiscent of a Black-and-white Warbler’s “squeaky wheel” song, except that in the guillemot it is slower and goes on much longer, sometimes for 30+ seconds, gradually rising and falling in pitch.  The equivalent sound in Pigeon Guillemot is much slower, alternating series of short “psip” notes with series of longer whistles.

Black Guillemot Seet Series
Black Guillemot Seet Series, Cape Bonavista, Newfoundland, 7/5/2007 (LNS 133369; click image to listen).

Pigeon Guillemot Seet Series
Pigeon Guillemot Seet Series, Carmel, California, 6/26/1961 (LNS 3856; click to listen. Spectrogram is from the fifth minute).

Black Guillemot’s vocalizations are much more likely to involve series of couplets with alternating notes, often in a long-short, long-short pattern.  Pigeon Guillemot’s vocalizations rarely involve long-short couplets for any continuing length of time; they are much more likely to contain the same kinds of notes repeated several times before switching to a new kind of note.  However, vocalizations in both species are highly plastic.

“Seer” calls

The most commonly heard call from both species of guillemot is a high-pitched, finely buzzy “Seer” that may recall a Cedar Waxwing.  These calls are rather plastic, but usually slightly downslurred.  They seem to average higher-pitched in Pigeon Guillemot, but are otherwise almost the same.

Black Guillemot
Black Guillemot “Seer” calls, Cape Pierce, Nova Scotia, 6/28/2007 (LNS 133380; click image to listen).

Pigeon Guillemot
Pigeon Guillemot “Seer” call, St. Lazaria Island, Alaska, 6/17/2007 (LNS 136454; click to listen).

Both species can also give shorter high-pitched notes, ranging from “seeps” or “pseeps: to quick sharp chips.

Even though birders in the field are not particularly likely to need to know the vocal differences between the guillemot species, I hope you find this information enlightening.  Personally, I find the sounds to be a fascinating glimpse into the evolution and taxonomy of these charismatic birds!


Ever heard of Pine Flycatcher?

Ever heard of Pine Flycatcher?

Pine Flycatcher - Durango, Mexico.  Copyright Andrew Spencer.
Pine Flycatcher, June 2015, Parque Natural Mexiquillo, Durango, Mexico. Copyright Andrew Spencer.

No group of birds in the ABA area is as infamous for being difficult to identify as Empidonax flycatchers. The very utterance of that dreaded name can make a beginning birder quiver in fear, and even the most studied of experts have a healthy respect for the challenge they represent. In earbirding terms, “Empids” are just about the ultimate argument for sound over sight. While they are no longer considered identifiable by voice only (knowledge of their visual identification has progressed by leaps and bounds the past few decades), vocals are still the best way to distinguish the various species.

The purpose of this blog post is to draw attention to one of the least known Empidonax species, one that has yet to appear in the ABA area, but is high on the list of many a birder as a potential vagrant: Pine Flycatcher (Empidonax affinis). An inhabitant of montane forests from just shy of the Arizona border to northern Central America, it favors (as one would expect) pine-dominated woodland.

While Pine Flycatcher has yet to appear north of the border, the species is already somewhat infamous for an “almost” that appeared in 2008-2009. For several weeks, an Empid in Choke Canyon, Texas, caused a good deal of excitement when it was identified as a Pine Flycatcher, both on visuals and vocals. Many listers visited and “ticked” the bird before it was re-identified as a Least Flycatcher. Even though, in the end, the bird wasn’t the hoped-for first ABA record, Pine Flycatcher was now “on the radar”.

Pine Flycatcher is still a species that could show up some day. As it is also one of the most poorly known Empid species, both visually and vocally, I’m hoping this blog post will help anyone trying to identify them.

Visually, Pine Flycatcher isn’t one of the more distinctive Empids. It tends to have an overall greenish cast to the upperparts and whitish/dingy coloration to the underparts, but it is not terribly colorful, even by Empid standards. Structure plays a better clue in its identification, as it has long primary projection (like Hammond’s), but also has a long, and narrow, bill with a completely pale mandible. This last feature can be especially useful, as the most common way to see a Pine Flycatcher (at least on the breeding grounds) is to be looking straight up at it as it perches high in a tree.

Away from the breeding grounds, the most commonly heard vocalization of any Empid is typically its contact call. In many species this vocalization is transliterated as a “whit”, while in others it sounds more like “pip”. Pine Flycatcher falls into the “whit” category, but with an important caveat. Almost all of the other species that have a “whit” call are at a similar pitch, and the difference between the calls are down to subtle differences in note shape that give the call a different quality. Pine Flycatcher is distinctly lower pitched, which gives its “whit” call almost a “pip-like” sound. It is very distinctive once learned, and would be a major piece of evidence in favor of a potential vagrant.Empid calls comp_m

In the graphic above (with thanks to Tayler Brooks for her help in its preparation), note how strongest (brightest) part of Pine Flycatcher’s call is nearly 1 kHz below the strongest part of the next lowest “whit”, that of Willow Flycatcher. It also has a smaller bandwidth (that is, covers a small range of frequencies), which contributes to the impression of a “pip” like quality. The other main factor determining the sound quality of each species’ call is the overall steepness of the note, with a larger rise over a shorter time imparting a “drier” (less musical) quality to the call.  On the far right is Hammond’s Flycatcher, just for comparison. Hammond’s give a true “pip”, the effect of having a call that is cleanly overslurred, without an overall rising pitch. This graphic is also a good one to study to visualize the differences between the western ABA breeding Empid species (the intricacies of their call notes will likely be the topic of a future Earbirding post).  You can listen to the various Empid call notes below:

The song of Pine Flycatcher is also distinctive. Though it is less likely to be heard in the US, it wouldn’t be entirely out of the question to have an overshooting bird set up territory somewhere like the Huachucas or Chiricahuas – a pair of Tufted Flycatchers nesting in Ramsey Canyon shows how the crazy isn’t entirely impossible! Like many other western Empid species, Pine Flycatcher has a song made up of three phrases, typically mixed together (especially at dawn). These are A) a bi-syllabic “PIT-wee”, B) a single syllabled “DIT”, and C) a burry “breee”. All of these phrases can be preceded by a variable number of quiet “pip” notes, especially phrase C, which nearly always has 3-4 “pips” included. The sonogram below has one of each phrase, given at a natural interval at dawn.  You can listen to the recording the sonogram is made from here.Pine Flycatcher song_m

As you would expect, Pine Flycatcher has a whole suite of other, more rarely heard calls. These tend to be analogous to the various calls given by our more familiar species, including a twitter series (the first recording linked below), a “pweep” call indicative of high agitation, superficially similar to calls made by Great Crested or Nutting’s Flycatchers (the second recording), and one in the vein of Dusky Flycatcher’s “du-hic” call (no publicly available recordings). These calls are unlikely to be heard from a vagrant bird, but should you find yourself standing in some towering pine forest in Mexico you would do well to keep an ear out for them!

Finally, there has been quite a bit of talk about a potential split of Pine Flycatcher involving differences in birds found north and south of the Isthmus of Tehuantepec.  The reasoning given is that the populations found on either side of the Isthmus have notably different songs.  While there aren’t a whole lot of recordings of the full song of birds found south of Oaxaca, the ones that I have been able to find (such as this one by Kathi Borgmann, and one kindly sent to me by Knut Eisermann) are disappointingly similar to the birds I’ve heard from further north.  While there do appear to be some minor differences in one of the phrases, it is perhaps more on par with differences among populations of Willow Flycatcher currently considered subspecies rather than the differences found between most full-fledged Empid species.  However, that is a conclusion drawn from only two recordings.  I’m quite willing to change my mind if someone goes out and gets a larger sample size showing a more profound difference!

The Swifts of Mexiquillo

The Swifts of Mexiquillo

The waterfall (cascada) in Parque Natural Mexiquillo, Durango, Mexico. Photo by Roberto González (CC-BY-2.0).
The waterfall (cascada) in Parque Natural Mexiquillo, Durango, Mexico. Photo by Roberto González (CC-BY-2.0).

We were excited to arrive at Parque Natural Mexiquillo before the break of dawn in June.  This underappreciated park is right on one of Mexico’s most celebrated birding routes, the Durango Highway — but it is an hour or two farther east than the Tufted Jay Preserve, and much less frequently visited.  The drier forests of Mexiquillo host a noticeably different avifauna than the wetter areas closer to the coast. Mexiquillo is more likely to produce birds typical of the dry “sky islands” of southern Arizona: Buff-breasted Flycatcher, Western Bluebird, Plumbeous Vireo, and even American Robin.  It’s much better habitat for the likes of “Type 6” Red Crossbills (which we recorded) and Eared Quetzals (which we did not find). Of course, it also has Red Warblers, Elegant Euphonias, and Rufous-capped Brush-Finches.

And it has a waterfall.

This is particularly noteworthy because the mountains of west Mexico are high and dry — and not much water means not many waterfalls.  Which means not many places to look for waterfall-loving swifts.

We were very interested to see what kinds of swifts would be flying around the waterfall at daybreak. A report from 2013 had us wondering whether this would be a good place to record Black Swifts, and a tantalizing recording from Chihuahua had us wondering whether something super-rare might be lurking under the cascada (and under the radar), like the “avian unicorn” White-fronted Swift (Cypseloides storeri).

When we first arrived, before first light, all was quiet, except for the roaring water and the aerial singing of Violet-green Swallows. Off in the distance, a Spotted Owl hooted. No swifts. Soon a dawn-singing Pine Flycatcher became a serious distraction. By the time the sun was fully up, we were totally engaged in all the other wonderful birds in the park, and we had completely given up on the swifts.  Until suddenly, something dark swooped by overhead with a sound like the crackling of electrical wires.  Chestnut-collared Swift!

Then another group of swifts flew by, and I heard a sound familiar from the mountains of Colorado: Black Swift.

For the next hour, these two species swirled over our heads, sometimes swinging pretty low. The encounter was simultaneously a letdown and a thrill.  On the one hand, it woke us up from the pipe dream of documenting some ultra-rare Cypseloides vocalizations. On the other hand, it was a terrific opportunity to spend time with two species that it can be awfully hard to see and hear well. It cleared up a long-standing mystery about whether the Black Swifts of Mexico might sound different from the ones farther north (answer: they don’t). And it gave us the field experience necessary to help identify the Chihuahua mystery swift as the first Chestnut-collared documented in that state.  One key takeaway: Black and Chestnut-collared Swifts can look astonishingly similar, even when you get good, prolonged looks.  But by sound, they can be separated right away.

While I concentrated more on the audio recording, Andrew was able to get some decent flight shots of the Black Swifts:

We were hoping to figure out whether the Black Swifts of Mexiquillo belonged to the northern subspecies niger, which breeds in the United States and Canada, or the southern subspecies costaricensis, which breeds in — you guessed it — Costa Rica.  The birds in Mexico have been attributed to each of these subspecies by different authors. The Mexiquillo Black Swifts were surprisingly well marked, some with bright white foreheads, some with extensive white scalloping below. So we thought this might be a mark for costaricensis. But when Andrew reviewed and photographed Black Swift specimens in the Harvard collection earlier this week, he found it nearly impossible to separate the subspecies by coloration. The size difference is noticeable on the specimen table, but is subtle enough that it would be virtually useless in the field.

So, the subspecies of the Durango birds remains a bit of a mystery. Any comments or discussion would be appreciated!

Mexico’s Mystery Owl

Mexico’s Mystery Owl

Juvenile Cinereous Owl, Rancho La Noria, Nayarit, Mexico, 4 June 2015. Photo by Andrew Spencer
Juvenile Cinereous Owl, Rancho La Noria, Nayarit, Mexico, 5 June 2015. Copyright Andrew Spencer

It’s not every day that you photograph and audio record a bird that has never been photographed or audio recorded before.

But that’s what we managed to do Thursday night, on the trails above Rancho La Noria in the state of Nayarit, Mexico. At the end of a madcap 12-day birding trip,  we finally connected with our primary target, one of the most elusive birds in Mexico – Cinereous Owl (Strix sartorii, also known as Mexican Barred Owl).

Known from about 30 specimens taken between 1873 and 1948, this bird has eluded almost all detection for the past half century, and its taxonomic status has been a major mystery.  Is it a subspecies of the Barred Owl of North America, as the American Ornithologists’ Union currently classifies it?  Is it a species in its own right,  as the International Ornithological Congress decided to call it following recommendations from a 2011 genetic study?  Or is it a subspecies of the Fulvous Owl, as some have speculated following recent records of Fulvous Owl from within the supposed range of Cinereous [1 2 3 4 5]?

To answer these questions, good photos and (especially) audio recordings of Cinereous Owl were sorely needed. And here are the first ones.

The search

When we first began considering a trip to West Mexico with our friend Carlos Sanchez earlier this spring, the Cinereous Owl was a major motivator, and we planned a good portion of the trip around it.  We contacted Steve N. G. Howell, the author of the field guide to Mexican birds, one of the only people to have encountered Cinereous Owl in the wild. He recommended the trails above Rancho La Noria (a.k.a. Cerro San Juan), where he had found the bird in the past, but had been unable to get audio recordings despite four trips specifically for that purpose. So we knew the bird wasn’t going to be easy.

Rather than head straight for Rancho La Noria, we started our search at Cerro La Bufa in Jalisco, on the recommendation that the habitat was very similar to that at Rancho La Noria, but more extensive, higher in quality, and easier to access by road. Nevertheless, in a night of owling that included playback of Fulvous, Barred, and Spotted Owls (on the assumption that Cinereous probably sounded enough like one of these to respond to a tape), we only managed to find Spotted Owl, Mottled Owl, and Whiskered Screech-Owl.

We tried again a few nights later on Volcan Nieve (a.k.a Nevado de Colima), the source of at least one of the specimens of Cinereous Owl. Again, no luck. However, in our limited time we were only able to survey a small portion of the habitat along one of several roads on the peak, so Cinereous Owl could easily be hiding up there somewhere.

We probably should have tried owling during our night at the Reserva Chara Pinta (Tufted Jay Preserve) near El Palmito, Sinaloa, but by that point in the trip we were too tired.  So our last chance for the mythical bird came down to our final hours in Mexico, at the spot Steve Howell had originally recommended.

The find

Carlos opted for hard-earned sleep instead of an exhausting wild goose chase, so we headed up the Rancho La Noria trails at dusk without him. Our process was the same as at the other sites: stop, listen for a couple minutes, and then try speculative playback of all three Strix owls, to see if we could get some kind of response. At our third stop, we didn’t even get to the playback, because we heard the bird we were listening for singing spontaneously far up the trail!  Just as Steve had described, it sounded a lot like Fulvous Owl, but slightly different.

Elated, we quickly moved up the trail to get closer to the still barely audible owl.  After several more stops it became apparent that we weren’t going to be able to get closer to it on the trail, so we tried some playback.  This did not succeed (at least initially) at bringing the singing bird closer, but we did start to hear what sounded like the wail call of a female Strix owl from the same area.  So we bit the bullet and started bushwhacking down a steep slope.  We didn’t make it far when the apparent male of our target pair called from right over our heads, twice!  Quick reflexes and the wonders of pre-record buffering, and we had the first ever recording of Strix sartorii (at least the first good ones, after our far more distant cuts a few minutes before).

And that was it.  After several minutes of silence we played back what we had just recorded, we played back Fulvous Owl, and we played back the female wail call of Spotted Owl – no response.  But we were still hearing a consistent wail call in the distance, and we reasoned that we would stand a better chance of getting more audio if we tried to get closer to the source of that call.  Two steep ridges later and we were far closer to the calling owl, and also on a trail that, had we but known of it, would have saved significant effort.  A bit more walking, and we were soon staring right at the source of the wail call – not a female as we had assumed, but a begging fledgling Cinereous Owl (see the photo above).

We spent the next hour or so with the young bird, documenting what we could, but the adult birds never showed.  So we decided to get a precious few hours of sleep and return before dawn, when we hoped the adults would vocalize more and possibly return to feed the juvenile.

A few hours later, this time with Carlos, we were standing on the same stretch of trail.  It didn’t take us long to find the same shrieking fledgling, but there was no sign of its parents.  As dawn approached and the adults never showed we got more and more nervous, but then finally the young one made a rather more urgent sounding call, flew off out of sight, and the sounds of a feeding event floated through the woods.

After two apparent feedings (both recorded; see above), an adult bird finally sang a hundred meters or so down the trail.  Some quick repositioning, a few audio cuts later, and a bit of playback, and we were looking right at it – an adult Cinereous Owl! Over the next half-hour we obtained all the evidence we could have hoped for, audio and visual.  A few other residents of the forest took some exception to the singing owl (including a White-throated Thrush and a Sharp-shinned Hawk), but in the end we had to tear ourselves away from it all.  It was time to return to Guadalajara, and the US, tired but thrilled at our success.

The analysis

While seeing such a rare animal was an awesome experience, the real interest for us was to see how these birds compared to both Barred and Fulvous Owls, and try to figure out where Cinereous Owl actually belongs.  All of the below is with the caveat that our sample size is small (just one adult, likely a male).


Most Strix owls, including Barred and Spotted, have two song types, one with notes in a species-specific rhythm, and another with notes in fairly steady series.  We recorded both song types from Cinereous.

At first glance, the rhythmic song of Cinereous Owl is far more like that of Fulvous Owl than the familiar “who cooks for you” of Barred Owl.  In fact, it is so similar that some might argue it is evidence that Cinereous and Fulvous could be considered conspecific.  But a closer look shows some differences.  The song of Cinereous is significantly lower pitched than that of Fulvous (which is to be expected since Cinereous is a larger bird).  And it consistently had two extra notes appended to the end, even when singing naturally before playback.  (This Fulvous Owl in Chiapas gave one extra note at the end of its song in response to playback.) The note shape and pace also differed, with Fulvous having more disyllabic notes given slightly faster, imparting a different cadence to the whole song.

The difference in the series song may actually be more striking, because it’s possible that Fulvous Owl lacks this vocalization type altogether. We have not been able to find a single recording of the series call from Fulvous Owl, even in their pair duets or in response to playback, when other species are more likely to give them.  Instead, excited Fulvous Owls seem to stick with the rhythmic song, often in duets, tossing in occasional “caterwauling” barks. Although it’s hard to prove a negative, it’s tempting to conclude that Fulvous Owl lacks a series song, given that good online recordings are available of over 15 different individuals/pairs calling, many in response to playback.

Since we weren’t able to find a recording of the series song from Fulvous Owl, we can only compare what we got from Cinereous with Barred Owl.  Compared to the series song from Barred, what we recorded from Cinereous was similar in pitch, but quite different in pace and overall pattern.  In Barred, the series song rises slightly but consistently at an even pace before ending in a slightly modulated, falling note.  In our sample of Cinereous, the series song started and ended at a lower pitch, with a few higher pitched notes in the middle, and a more stuttered pace throughout.


Plumage-wise, Cinereous Owl clearly looks more like Barred.  It isn’t hard to see why they were lumped for so long, especially considering that the vocal differences were unknown.  But that’s not to say they’re identical, and Cinereous Owl can be told from both Barred and Fulvous Owls with a good view.

The first thing that immediately struck all three of us when we saw the adult Cinereous Owl was how pale its face was.  Compared to both Barred and Fulvous Owls, the facial disk of Cinereous Owl is paler (without obvious darker concentric rings), and compared to Fulvous, grayer.  Indeed, the overall colder gray coloration is a good way to distinguish Cinereous from Fulvous Owl.  This is most obvious in the differences of the underparts pattern (dark sepia streaks on an off-white background for Cinereous vs. warm brown streaks on a light buffy background for Fulvous), and in the color of the collar (cold gray for Cinereous vs. golden brown for Fulvous).  Compared to Barred, which is far more similar to Cinereous in these features, the main difference is that the streaks below on Cinereous are slightly darker, not a medium brown.  See the photo below for a comparison of all three species (and click to see a larger version without the labels).  Here also are a few more good photos of Fulvous Owl from Oaxaca [1 2 3 4] and from farther south [1 2 3 4 5].

Left: Barred Owl, Tampa, FL, 23 February 2014. Photo by Trish Hartmann (CC-BY-2.0). Center: Cinereous Owl, Rancho La Noria, Nayarit, Mexico, 5 June 2015. Copyright Andrew Spencer. Right: Fulvous Owl, Cerro San Felipe, Oaxaca, Mexico, 18 March 2015. Copyright Andrew Spencer.

A final, if far more subtle visual clue, deals with the feathering on the toes.  In both Barred and Cinereous Owls, the feathers extend all the way down the toe to the base of the claw.  In Fulvous, the toes are partly bare of feathering.  This can be difficult to discern in the field, but with good photographs it can be a useful feature (and it is visible in the large version of the comparison photo, which you can see by clicking on the photo above).

Conclusions and Questions

1) While our sample size is limited, what we were able to audio record clearly supports considering Cinereous Owl separate from Barred Owl.  Vocally, the two are not even close. Add that to the genetics in the 2011 paper, and there’s no reason to continue considering sartorii a subspecies of Barred Owl.  (This is also why we support the name “Cinereous Owl” instead of “Mexican Barred Owl”).

2) Is sartorii perhaps a subspecies of Fulvous Owl? This situation is a bit more murky.  Vocally, the two are close, at least in the rhythmic song. But with subtle differences in that song, and more marked differences in plumage and size, we think the case can be made that both be considered separate.  If it can be determined definitively that Fulvous Owls lack a series-type song, that would be a strong argument in favor of separate species status for Cinereous.

3) Several people have wondered whether the apparent Fulvous Owls recently found in Oaxaca are really Fulvous Owls. We think our photos (in conjunction with the photos of Fulvous Owl linked above, and the many available audio recordings) demonstrate that the Oaxaca birds can be definitively identified as Fulvous Owls, and ruled out as Cinereous. But putting this question to rest raises a host of others. How far north and west are Fulvous Owls found? Do they currently overlap with Cinereous Owls in Oaxaca (which has specimen records for both)?  Is one population expanding?  Are the two interbreeding?  Do intergrades exist?

What is needed now are more recordings and photos, especially from the other parts of Cinereous Owl’s range, in Guerrero and (even more importantly) Veracruz and Oaxaca.  Hopefully the recordings we were able to obtain from Nayarit will help anyone out there trying!

Further reading

For more information on Cinereous and Fulvous Owls, check out the detailed bibliography at the end of Michael Retter’s blog post from 2012. At this writing, most of the links are broken if you try to click on them, but will work if you copy and paste the URLs into your browser.

How to Identify a Timberline Sparrow

How to Identify a Timberline Sparrow

Timberline Sparrow - Andrew Spencer
Timberline Sparrow, Mount McIntyre, Yukon. This individual was an especially boldly marked and patterned individual. Copyright Andrew Spencer.

Brewer's Sparrow - Luke Seitz
Brewer’s Sparrow, Mojave Desert, California. Copyright Luke Seitz.

It’s hard to deny – among many birders there is a preconception that sparrows are brown and boring looking.  And while I would argue that point for many of the species in the group, it isn’t very easy to defend Brewer’s Sparrow (Spizella breweri).  It’s a small, brown, and subtly marked bird that doesn’t stand out much when you see it.  But just wait until you are standing on a sage flat in the pre-dawn and hear a Brewer’s Sparrow!  There are few auditory experiences in the west as sublime as being out in the crisp, cool air, surrounded by the smell of sagebrush, and hearing the tinkling buzzes of Brewer’s Sparrow all around you.

In addition to having one of the best voices in the west, Brewer’s Sparrow also has an interesting taxonomic facet to its name.  The nominate breweri subspecies is familiar to most anyone who birds in the appropriate habitat.  Less well known is that there is another subspecies – taverneri – that was first described as a separate species from far to the north of most Brewer’s Sparrows (Swarth and Brooks 1925).  Unlike the nominate subspecies, it is restricted to stunted, willow-dominate thickets right below treeline – hence the common name of “Timberline” Sparrow.  It was later lumped with Brewer’s Sparrow, but recent morphometric and genetic studies have again raised interest in the form (see Klicka et al 1999). For a long while conventional wisdom said that Timberline Sparrow occurred only well north of Brewer’s, in an isolated range centered around southern Yukon.  But as more people studied the bird it was found to also occur near treeline further south, first in the Canadian Rockies, and then in Montana.  Birds have been found in similar habitats even further south, to at least southern Colorado, but it is less well known what to make of these birds (more on this later in the post).

Timberline Sparrow - Andrew Spencer
Timberline Sparrow, Jasper National Park, Alberta. This individual (a female) was among the most dully patterned individuals I saw, closest to Brewer’s in appearance. Copyright Andrew Spencer.

Before I dive into the vocals of these birds, a bit on the visual identification.  As Brewer’s Sparrow doesn’t have much in the way of obvious field marks under the best of circumstances, it shouldn’t come as any surprise that telling Timberline from the nominate isn’t easy by any means.  The most often cited field marks are usually bill size (Timberline has a longer but thinner bill), strength of the face pattern and crown streaking (noticeably stronger in Timberline), overall color (noticeably grayer in Timberline), and sometimes some faint streaking on the flanks or sides of the chest in Timberline.  All of these marks are hard to use in isolation, at least for someone without a lot of experience with both forms.  Personally, I believe that if I spent time closely looking at migrant or wintering Brewer’s Sparrows I might be able to find birds that I have a strong suspicion are Timberlines, but that I wouldn’t be able to make a claim with certainty.

Enter vocals.  Differences in the song between Brewer’s and Timberline are the most often cited difference between the two.  Before getting into the nitty-gritty of it all, though, we need a bit of a primer in the sounds that Brewer’s Sparrow (sensu lato) make – it’s more complicated than just song and call.  That wonderful dawn chorus I mentioned in the intro to this post is what is known as the “long song” – a successive series of different trills, typically starting out very high-pitched and then descending into a long string that can sometimes last 20 seconds or more.  Mostly given very early in the morning and late in the evening, it can sometimes be heard at other times of the day but generally less often.  Below is an spectrograph of long song from Mesa County, Colorado, to demonstrate the structure of the voc.Brewer's Sparrow_long song spec

The other song of the species is appropriately known as the “short song”.  This vocalization, much shorter than long song, is typically comprised of one to three different trilled phrases, each phrase most often (but not always) even in pitch, but different in pitch from the previous one.  It is also one of the most variable songs in the west – some birds can sound startlingly like Chipping Sparrow, while others incorporate elements almost reminiscent of Prairie or Blue-winged Warbler songs. When you hear a Brewer’s Sparrow during daylight hours it is much more likely to be giving this vocalization than the long song, though again there are exceptions to this.  Most Brewer’s Sparrows only have one short song that they give, whereas long song can be variable in terms of which trilled elements they include.  If you want to read more on the different song types, Arch McCallum has a page dedicated to the topic here.

Like most other sparrows, Brewer’s also has various calls – flight call, chip call, twitter, etc.  I don’t really go into those here, in part due to very small sample size available from Timberline Sparrow of these calls, and also since I don’t personally believe that they can be identified to form.

Of the two song types, it is long song that I believe is most readily distinguishable between the two forms.  The “classic” long song of breweri typically sounds very buzzy throughout, descending from very high-pitched, insect-like buzzes to lower pitched buzzes, but rarely having anything that one would describe as “musical” about them.  This is due to the broadband nature of all of the elements of the song.  Most phrases in the long song are made up of two alternating notes that comprise the buzz, and in nominate breweri both of these elements are broadband, or if one is less so then the broadband one dominates.  Below are several examples of long song from breweri from around their range.

The long song of Timberline Sparrow, on the other hand, gives a much less buzzy impression.  The overall structure of the song is the same – it typically starts out with very high-pitched, insect-like phrases that quickly descend into the variable series of trills.  The difference comes in the fact that elements of the majority of the phrases in the song are noticeably less broadband than in breweri, so that they sound more musical and tinkling.  This difference isn’t huge, and there is considerable variation, so it is important to listen to enough of the song to get a general impression rather than make the call on just one or two elements.  In general, though, I think that once you’ve gotten a handle on how each subspecies sounds when long singing you can tell the different most if not all of the time.  Listen to several examples of Timberline Sparrow below (from four different birds; the first three are from the same individual in the Yukon, the next five from different birds in Alberta) and see if you can tell the difference:

When it comes to short songs, these differences can also apply.  Unfortunately, short songs seem to be a less reliable way to distinguish Timberline from Brewer’s.  The main reason for this is probably that in long song there are lots of difference elements and the impression the entire vocalization gives is a composite of how they sound, whereas in short song you only have a few to draw on.  So say you get a Brewer’s on the musical end of the range of variation and a Timberline on the buzzy end there could be some overlap.  That said, I do believe that on average the short songs of Timberline are less broadband and more musical.  First, here are a variety of breweri short song, both on the buzzy and musical ends of variation:

And here are two examples of Timberline short songs, one on the musical end and one on the buzzier end:

So by now you’re thinking “great! Now I can easily tell the difference between Timberline and Brewer’s, at least by the long song”).  Unfortunately there’s reason to be cautious; the differences between these two forms aren’t entirely black and white.  There are some recordings out there of breweri that are less buzzy than average, approaching Timberline.  And when you listen to the long songs of Brewer’s that I link above you can hear that the examples from California seem to be a bit less buzzy than those from Colorado.  When you combine this with a relatively small sample size of available Timberline Sparrow recordings and it’s hard to be completely certain how reliable the vocal differences are.  That said, I personally have never heard a Brewer’s Sparrow that sang a long song entirely like that of the Timberline Sparrows I’ve heard.  But if you’re investigating a suspect bird on territory out of the known range of one form or the other, I would want more than just one line of evidence before making the call.

This is especially relevant to birders in states where Brewer’s Sparrows have been found breeding near or at treeline but south of the known range of Timberline Sparrow.  A few years ago there was a lot of buzz about this issue in Colorado, where I am from.  Most birders in the state accepted that these birds breeding in highland stunted willow patches were a southern range extension of Timberline Sparrow, and that they were different than the more widespread Brewer’s Sparrow they were all familiar with.  And I’ll admit I was also in that camp, at least to start.

But then back in 2007 I actively tracked down two different populations of these highland birds.  I was rather surprised, at the time, at how un-different they were.  The first group I recorded all sang songs that sounded unremarkable to me.  In the second group, though, I did find a bird that was noticeably more musical and stood out, but it was surrounded by birds that were “normal”.  Here’s the recording I made of that more musical sounding individual.  Sounds a lot like Timberline Sparrow, doesn’t it?

What was going on?  To this day I am not entirely sure.  At the time my theory was that lowland breeding breweri moved upslope after their first brood to breed in dry stunted willow habitat superficially similar to the sage flats they used lower down.  But now I am unsure.  I didn’t record any long songs from either of those populations, and knowing what I do now after having more experience with definite Timberline Sparrows I am unwilling to make a call until I record some long songs from these southern highland birds.  It remains one of those exciting areas that needs to be looked into, a place where anyone with a recorder willing to make the effort to be standing on a high-mountain hillside waiting for the dawn can help unravel a mystery.

Common vs. Chihuahuan Ravens

Common vs. Chihuahuan Ravens

Common Raven, Bryce Canyon National Park, Utah. Photo by National Park Service (public domain).
Common Raven, Bryce Canyon National Park, Utah. Photo by National Park Service (public domain).

Every once in a while, somebody will put together a list of underappreciated identification challenges among North American birds.  Common and Chihuahuan Ravens always make the list.

Here in Colorado, we have lots of ravens. Conventional wisdom says the ones on the southeastern plains are Chihuahuans, and the rest are Commons. But not everyone accepts the conventional wisdom. Some believe that Chihuahuan Ravens are not only the default raven in the southeast, but also regular wanderers north to the Wyoming border.  Others maintain that to find a real Chihuahuan Raven, you have to go south of Colorado altogether.

For a long time, I haven’t been sure what to believe.  But I’ve been browsing my own audio collections and those online, and I’m starting to think that voice is actually a very good character — maybe the best field character — for separating these two species.  I’m willing to be convinced otherwise, and I’m interested in hearing from anybody who has experience with these two birds.  But at the moment, here’s what I think I know about the sounds.

Chihuahuan Raven

It seems to me that Chihuahuan Raven may actually have one of the most predictable (least variable) voices of any North American corvid.  Its classic call is a harsh “caw” that is deeper and much more coarsely burry than most “caws” of American Crow.  It is of medium pitch, slightly nasal, and barely upslurred.  It actually has something of a quacking quality — reminiscent of the harsh-sounding quacks that female Mallards do when they’re agitated.

Once you start comparing multiple recordings, this call of Chihuahuan Raven starts to seem remarkably consistent rangewide and from bird to bird:

Occasionally a couple of the birds in these recordings switch to some similar but slightly higher-pitched calls.  And females will also do the clucking gurgles that many female corvids do (you can hear some on my recording above from Chihuahua).  Other than that, though, these six examples are all so similar that they could almost be from the same individual bird.

Common Raven

From one of the least variable North American corvids, we transition to one of the most variable.

I’ve seen a number of sources argue that Chihuahuan Raven averages higher-pitched than Common Raven. This is a really problematic statement. It’s true that if you hear any really low croaks, you can be sure they’re from a Common:

But if you hear any really high croaks, you know they’re also from a Common:

(The ravens in that Red Top Ranch recording, by the way, were flying in a loose flock of 5 over featureless shortgrass prairie in southeast Colorado — location, habitat, and behavior all suggesting Chihuahuan Raven. But the voice doesn’t seem to match.)

Common Raven runs the gamut of possible pitches and tone qualities, from far below Chihuahuan’s pitch range to far above it. Chihuahuan occupies a limited space in the middle. Not only are Commons variable, but they often change sounds while you’re listening to them, unlike Chihuahuans. I might go so far as to say that if you hear a group of ravens giving a whole bunch of different-sounding calls, you’re almost certainly listening to Commons:

Some recordings of Common Raven come pretty close to matching the typical call of Chihuahuan Raven, but exact matches seem to be rare.  Interestingly, the most Chihuahuan-sounding Commons I’ve found are mostly from California:

You may recall that over a decade ago, Common Ravens in California were found to be genetically distinct from other Common Raven populations — more closely related to Chihuahuan Ravens than to other Commons. So should we split Common Raven?  Or reclassify California’s ravens as Chihuahuan?

Well, maybe not.  The study found that Chihuahuan Raven and “California” Common Raven were still pretty distinct genetically. And more recent followup research has showed that “northern” Common Ravens and “California” Common Ravens have apparently been interbreeding extensively in western North America for a very long time, and the genetic differences between them may be decreasing over time.  You could make a case for lumping ALL of North America’s ravens into one species, but I’m not sure that’s warranted, especially if Chihuahuans and Commons are not regularly interbreeding. More studies will need to be done on that.

On the whole, I could be underestimating the vocal variation in Chihuahuan Raven. And maybe there’s a lot of interbreeding going on in places like southeast Colorado. But at the moment, it really looks to me like Chihuahuan is a very predictable-sounding bird.  And if it is, then vocalizations may be the single best way to identify one in the field.

A Bicknell’s Thrush Critique

A Bicknell’s Thrush Critique

{We interrupt our series on describing bird sounds to bring you this special post.}

In 1995, in the 40th Supplement to their checklist, the American Ornithologists’ Union recognized Bicknell’s Thrush (Catharus bicknelli) as a full species, splitting it from the Gray-cheeked Thrush (Catharus minimus) on the basis of “differences in morphology, vocalizations, habitat preferences, and migration patterns.”  In support of these differences, they cited two papers: Ouellet 1993 and Evans 1994.

The split came under fire earlier this year in an extensive Xeno-Canto forum discussion that focused mostly on the vocal evidence.  Dan Lane gave a rather scathing assessment of Ouellet’s 1993 paper; Andrew Spencer testified that both Bicknell’s and Gray-cheeked respond to playback of one another’s songs, in contradiction of one of Ouellet’s key claims; and I made comments critical of Evans’ 1994 paper, which described differences in the nocturnal flight calls.

Recently, the conversation was joined by Bill Evans himself, the author of the 1994 paper.  In case you don’t know, Bill Evans is one of the great bird investigators of our age — one of the prime movers behind the last few decades’ resurgence in the study of nocturnal migration.   In his Xeno-Canto comment, Bill presented a detailed defense of his paper.  I promised to respond when I got a chance.

I have great respect for Bill Evans and what he’s taught all of us about birds and their sounds, but I’d like to reassess the evidence and arguments in his 1994 paper.  In my opinion, that paper simply did not present strong evidence for a consistent difference in flight calls between Bicknell’s and Gray-cheeked Thrushes.  And yet many have cited it as though it did, not least the AOU when it split the two species.  I’d like to get people thinking about this paper more critically.  Hence this blog post.

Evans’ original case

You can read Evans’ paper online for yourself, but I’ll summarize it here in a nutshell.

  1. Bill Evans collected a number of recordings of nocturnal flight calls of apparent Bicknell’s/Gray-cheeked Thrushes, from Minnesota, Alabama, west-central New York state, and Florida.
  2. He noticed that the Florida calls tended to have a much earlier peak and a higher maximum frequency than the calls from the other three locations (with little to no overlap, according to his table).  In other words, he detected two discrete types of “Gray-cheeked-like” flight calls.
  3. He argued that Bicknell’s Thrush would be expected to migrate directly through Florida, but that only “regular” Gray-cheeks were likely in the other three locations.
  4. He found a daytime recording of a Bicknell’s from Mount Mansfield in Vermont that was a close match for one of his nocturnal Florida calls, and a daytime recording of a Gray-cheeked from Manitoba that closely matched one of his nocturnal calls from outside Florida.  Here’s the figure he used to illustrate this point:

And by virtue of the evidence above, he argued that the nocturnal flight call of Bicknell’s Thrush differs consistently from that of Gray-cheeked Thrush.

When I first read this paper, I found it reasonably convincing, mostly because of Figure 1.  The top-to-bottom similarities and left-to-right differences are obvious, and they tell a clear story.  It’s an excellent piece of visual rhetoric (and I say that admiringly, as a teacher of scientific writing and rhetoric).

But as I started to research the vocal repertoires of Bicknell’s and Gray-cheeked Thrushes for my field guide project, I realized that Figure 1 is far too simple, far too neat. The daytime calls of Gray-cheeked and Bicknell’s Thrushes are extremely variable. So variable, in fact, that if you pick the right recordings, you can construct an alternate version of Figure 1 from Evans 1994, with the two species switched:

Bill said both in his 1994 paper and again in his Xeno-Canto comment that he couldn’t find any Gray-cheeked daytime calls that matched his purported nocturnal Bicknell’s, and he couldn’t find any Bicknell’s daytime calls that matched his purported nocturnal Gray-cheekeds.  Well, these look pretty darn close to me.  The Gray-cheeked call at lower left is a pretty good match for the early-peaked shape of the purported Bicknell’s at upper left, and it’s got almost exactly the same peak frequency.  It’s from the exact same Churchill, Manitoba recording as the one Evans used to create the daytime Gray-cheeked spectrogram “D” (the high, early-peaked call occurs at 1:20, while the “D” call is either the one at 1:26 or the one at 1:36).

Meanwhile, the call at lower right is from one of Andrew Spencer’s recordings of Bicknell’s Thrush from New Hampshire.  Note the “buffalo-humped” shape and the peak frequency all the way down at 4 kHz.

So both species can give high-frequency, early-peaked calls during the day.  And both species can give low-frequency, late-peaked calls during the day.  So why couldn’t they both give both types of calls at night?

The migration-route argument

Evans 1994 argues that Bicknell’s would be expected to migrate through Florida at the time of his recordings, but never discusses whether Gray-cheeked Thrush might also pass through at that time.  But Gray-cheeked appears to be more common than Bicknell’s as a migrant in Florida.  A 2005 paper by Glen Woolfenden and Jon Greenlaw found that of 54 Florida specimens, 47 were Gray-cheeked Thrush, and only 4 were Bicknell’s (with 3 remaining unidentified). Eleven of the Gray-cheeked specimens were from eastern coastal counties, and Woolfenden and Greenlaw could find no differences in migration dates.  A similar situation seems to exist in other southeastern states: Lee (1995) re-examined 24 specimens taken in North Carolina and found that 23 were Gray-cheeked and only 1 was Bicknell’s.

The sheer number of Gray-cheeked specimens from Florida (even eastern Florida in May) suggests that the peninsula is a regular migration route for at least part of the population.  And the ratio of southeastern US specimens suggests that in migration, Gray-cheeks outnumber Bicknell’s throughout the region.  Gray-cheeked winters in northern South America from Columbia east to Guyana — largely south and east of Bicknell’s wintering range on Hispaniola — and Gray-cheeked is reportedly a “trans-Gulf migrant” (according to BNA, etc.).  All of this suggests to me that any given Gray-cheeked-or-Bicknell’s flight call recorded in Florida is more likely to be from a Gray-cheeked.

Evans 1994 reports only high-pitched, early-peaking flight calls from Florida, but Bill’s Xeno-Canto comment indicates that subsequent sampling there has turned up low-pitched, late-peaking flight calls as well.  He wrote, “you don’t get a regular stream of “buffalo humped” GCTH calls in the mid-eastern Florida coast in May unless you have a sustained period (typically 2 days or more) of westerly winds. And there is nowhere else I’ve found in eastern US where one can record a pure set of the higher pitched, steadily descending GCTH calls like those I’ve recorded from eastern FL in May.”

This is an interesting claim.  I’d expect Gray-cheeked Thrushes breeding in Quebec and Newfoundland to pass through Florida even when winds weren’t westerly.  The situation Bill describes is consistent with a scenario in which the high-pitched, early-peaking flight calls are given by eastern Gray-cheeks, while the low-pitched, late-peaking flight call is the hallmark of the western Gray-cheek, with Bicknell’s either unrepresented in the sample, or overlapping eastern Gray-cheeks. At the very least, I don’t know how Bill can rule such a scenario out.

Also, the phrases “a regular stream” and “a pure set” ring my alarm bells.  They suggest that “irregular streams” and “impure sets” — such as the odd “Gray-cheeked-type” call on a non-westerly Florida wind, or the odd “Bicknell’s-type” call outside the range of Bicknell’s — might be dismissed as “atypical,” resulting in unwitting confirmation bias.  (More on the dangers of the word “atypical” here.)

Daytime call variation

I’ve put together two animated spectrogram GIFs to show how the daytime calls vary in individual Gray-cheeked and Bicknell’s Thrushes.  They both loop at 10 frames per second.

65 consecutive calls from an individual Gray-cheeked Thrush, Nome, Alaska, 6/1/2007. Recording by Gerrit Vyn. Click to listen to original (Macaulay Library catalog #137552).

67 mostly-consecutive calls from an individual Bicknell’s Thrush, Jefferson Notch, NH, 6/24/2008. Recording by Andrew Spencer. Click to hear original (split into 4 Xeno-Canto files).

Both GIFs are at exactly the same scale: the top of the graph is 6 kHz. Notice how peak frequency changes, as well as call shape.  No call forms are exactly shared, but some are strikingly similar.

Notice, too, that the calls don’t vary at random.  Birds of both species repeat the same call many times in a row. Then they either switch immediately to a different call type, or transition into the new type via 2-3 intermediate calls.  The Gray-cheeked uses four different call types in this sample; the Bicknell’s uses five.  If we had longer recordings, it’s likely we’d hear more call types from each individual: recordings and various published reports put individual call repertoires at up to 10 in both species (e.g., Ball 2000).  And repertoires of neighboring birds tend to be similar, as evidenced by recordings and reports of call matching by neighboring birds.  But across the geographic range, these calls vary greatly; Marshall (2001) found a huge variety of call forms across the continent.

These locally-similar but regionally-different repertoires of call notes vary in a pattern like that seen in Red-winged Blackbirds and Lapland Longspurs, and the pattern strongly suggests that the burry daytime calls of Gray-cheeked and Bicknell’s Thrushes are learned, not innate.  If the calls are learned, then regional differences would be expected to arise over time, in exactly the same manner as song dialects.  Some call types might be heard less often in one population and more often in another; some might become exclusive to a particular area.  Gray-cheekeds breeding in, say, Quebec or Newfoundland (and migrating through Florida) might use higher-pitched, earlier-peaking calls with greater frequency than Gray-cheekeds breeding farther west.

And this raises several questions.  If each individual thrush knows 5-10 different daytime versions of its burry calls, which one (or ones) does it give during nocturnal migration?  Or is the night flight call completely separate from the daytime calls?  Is it innate or learned?  Do individual birds have repertoires of night flight calls, or is there just one per bird?  When we see variation in nocturnal calls, how much of it is individual, how much of it is dialectal (that is, geographic), how much of it is repertoireal (I just made up that word), and how much of it is due to mere plasticity?

These questions matter, and right now we don’t know the answer to any of them.  Even if we know the answers to these questions for other species, it may not be safe to assume that Gray-cheeked and Bicknell’s are the same.  After all, among North American Catharus thrushes, only Veery shares their trait of having not just one daytime flight-call-like sound per individual, but a repertoire of such calls.


Based on all the research I’ve done, I have a hunch that during the day, Bicknell’s is indeed more likely on average to give the higher-pitched, early-peaking calls, while Gray-cheeked is indeed more likely to give the lower-pitched, later-peaking calls.  But the overlap seems pretty much complete, and I suspect that few, if any, nocturnal call forms are diagnostic for one species or the other.  Remember, in bird identification, it’s not enough to find a match; you have to rule out other potential matches.  If Gray-cheeks are capable of making the Florida-type calls, then how do we know they didn’t?  How do we know the higher-pitched, early-peaking calls aren’t more common in eastern populations of Gray-cheeked?  How do we know that the existence of two call forms in Florida isn’t an artefact of limited sampling?

I’m perfectly willing to believe that Bicknell’s and Gray-cheeked Thrushes do differ in nocturnal flight calls.  I’m even willing to believe there’s little overlap.  But if I’m going to believe it, I need some evidence more solid than anything I’ve seen yet.

Long Calls of Gulls

Long Calls of Gulls

Herring Gull in “oblique” posture at end of long call, Acadia National Park, Maine, 8/3/2008. Photo by Dick Daniels (CC 3.0)

A couple weeks ago, I was puzzling over a group of distant gulls on a reservoir in Wyoming.  I was fairly sure they were all the same species, but through my binoculars, I couldn’t quite decide whether I was looking at Ring-billed or California Gulls.  Then, from across the water, I heard one of them give a hoarse, slow “koo-WEEE? kweee? kweee? … QUIRK! …  QUIRK!” and I confidently checked the “Ring-billed” box in my checklist app.

As I wrote last year, most people don’t listen to gulls much.  But as I’ve paid more attention to them over the past year, I’ve realized that many species can indeed be identified by sound alone, and this fact has greatly improved my birding skills.  In today’s post I’m hoping to provide a basic framework for beginning to identify some of the species by sound.

What’s a “long call”?

The “long call” is the most complex vocalization in a gull’s repertoire, and one of the most frequently given. “Long calls” can be heard year-round in a variety of situations, but they most often serve as aggressive signals directed at other gulls.  They vary within individuals based on excitement level, but individuals may be able to recognize each other by long call.  The long calls of different species tend to sound rather different.

You can think of the archetypal “long call” as containing three parts:

Ring-billed Gull long call, Washington County, Colorado, 9/7/2009
  1. The intro notes are highly variable depending on the bird’s level of excitement, and are often totally absent.  When present, they usually take the form of short barks, as in the example above; long mewing wails; or a mix of the two.
  2. The squeals are the start of the long call proper.  They are by far the loudest, and usually the longest and highest-pitched, notes in the call, and they usually coincide with a distinctive motion of the head, such as a deep bow and/or a backward head toss.  There are usually only 1-2 squeals, but occasionally three or even more if a bird is excited.  (Some species don’t really “squeal” at this point in the call, but I couldn’t think of another good name for these notes.)
  3. The terminal series is the series of similar notes that ends the call.  The  number of notes varies with excitement level, but the sound and speed of the notes tend to be fairly consistent within species, making this generally the most useful part of the call in identification.  Most species adopt an “oblique” posture for the duration of the terminal series (see Herring Gull photo above), but some, like Ring-billed Gull, give distinctive motions of the head with each note.

Most large North American gulls sort out pretty well into groups that follow similar patterns in their long calls.

  • High yelpers, usually clear: Herring, Glaucous, Western, Glaucous-winged
  • Low yelpers, often hoarse: California, Lesser Black-backed, Great Black-backed
  • Slow squealers: Ring-billed and Mew
  • Nasal: Heermann’s, Franklin’s, Laughing

The High Yelpers

These are the “classic seagull” sounds as brought to you by Hollywood.  In this group, the terminal series usually consists of clear (not hoarse) high-pitched yelps, often without much change in pitch or speed. Western and Glaucous-winged average lower in pitch than the other two species, and tend to have simpler calls, very often omitting the intro notes and squeals.  Glaucous and Glaucous-winged average slower than Herring and Western, with fewer notes per call.  However, individual variation often makes it difficult to tell these four species apart by vocals alone.

The Low Yelpers

This group resembles the first group, except that the pitch of the terminal series averages lower and more nasal (almost more “bugling” than “yelping”), and there is a much greater tendency towards hoarse or harsh tone qualities.  Great Black-backed tends to be the lowest of these.

The Slow Squealers

Ring-billed Gull long call, Adams County, Colorado, 1/24/2013

Mew Gull long call, Alaska, 6/28/2006. ML catalog #132251 Click here to listen

These two closely related species have a distinctively slow delivery — typically about two notes per second in the terminal series.  The pitch tends to be high and the quality rather squealing.  Tone quality separates them: Ring-billed’s voice is distinctively scratchy or screechy, especially in the terminal series, while Mew tends to have a much clearer tone.  In Ring-billed, at least, each note in the terminal series is usually accompanied by a distinctive vertical pumping of the neck with the bill held skyward, as though the bird were jabbing its bill at an airborne opponent with each cry.

The Nasal Gulls

Laughing and Franklin’s Gulls were recently moved from the genus Larus into the genus Leucophaeus, reflecting our understanding that they are not closely related to the other gulls on this page.  Their immediately distinctive voices, high-pitched and nasal, are evidence of this evolutionary distance.  Their long calls are also unique in structure, lacking the “intro>squeal>series” pattern of other gulls.  Laughing Gull is the only gull that consistently starts with a fast series of notes followed by a slow series of notes. Franklin’s, meanwhile, usually gives a simple slow series of rising notes at one speed.  The fast notes of Franklin’s, if they occur at all, usually come at the end.

Then there’s Heermann’s Gull.  There’s no mistaking it for any other West Coast species.  Imagine the long call of a Herring Gull, as performed by a Red-breasted Nuthatch:

There’s a lot more to gull vocalizations, but hopefully this post can get people started on what to listen for.  I still have a lot to learn about identifying these birds by sound, and I’d be glad to get your insights in the comments.

Common and Hoary Redpolls

Common and Hoary Redpolls

Apparent Hoary Redpoll, eastern Ontario, 2/4/2009. Photo by Seabrooke Leckie (CC 2.0).

Last winter, in the Great Redpoll Invasion of 2012-2013, huge numbers of these little Arctic birds pushed farther south than anytime in recent memory, prompting a surge of interest in redpoll identification and taxonomy.  My good friend Andy Boyce even got people talking about whether Hoary and Common Redpolls are separate species, or identifiable forms, and the many resulting discussions taught me (and others, I’m sure) a lot about how we define and detect species boundaries.  And the conversation hasn’t quit: Tom Johnson and Luke Seitz are contributing with a redpoll identification article in the July/August issue of Birding.

During the course of these discussions, a couple people argued that Common and Hoary Redpolls have been shown to differ in vocalizations.  To verify this claim, I set out to track down the relevant peer-reviewed literature, including a paper in Swedish, two key publications in Russian, and a book in German.  It took some six months, but thanks to the help of a terrific interlibrary loan crew (and some assistance from Albert Lastukhin in Russia), I finally managed to get copies of all the sources.  My dusty old fluency in Russian came in handy, as did my one year of college German.  (Oh, and Google Translate.)

The original claims of differences in the literature

In 1981, Maria Zablotskaya provided a detailed spectrographic and behavioral analysis of the vocalizations of Common Redpoll; the following year, she teamed up with Boris Veprintsev to do the same for Hoary Redpoll.  These authors are often cited as having found differences in the sounds of the two species, but it’s important to note that they did not explicitly address the issue of species differences (or species boundaries) at all.  The first paper treated Common Redpoll only; the second paper discussed Hoary Redpoll only, with only a few passing comparisons to Common.  They named several more calls for Common than they did for Hoary, but this is likely due to the fact that they spent less time with Hoaries and had fewer recordings available for analysis.  A careful reading of both papers (along with Ernst 1998) shows that there probably aren’t any sounds in the repertoire of one species that don’t have an equivalent in the other species.  This doesn’t necessarily mean that the species sound the same, but it does mean that they give similar types of calls.

In the face of doubt about whether the two redpoll species should be lumped, a couple of authors stepped forward in the mid- to late 1980s to defend the two-species concept.  Ulf Molau’s (1985) “The redpoll complex in Sweden” focused on plumage and measurements, but also made the following important claims about vocalizations (translated from the Swedish by me, with help from Google Translate):

Common and Hoary Redpoll differ not only in size and plumage characters, but also in vocalizations. The low-key contact sounds uttered on the ground or in a tree crown (or in a cage) may in Common Redpoll be described as a catchy “chit-chit-chit.” Hoary Redpoll’s contact vocalization is a rougher, almost House Sparrow-like “chirp-chirp.” Similar differences also exist in the familiar “flight song” which is rougher and more metallic in Arctic Redpoll. Both Common Redpoll races flammea and cabaret are, however, completely identical as far as vocalizations are concerned.

These conclusions are based on Molau’s own observations of wild, banded, and long-term captive birds.  He did not provide spectrograms.

Marc Herremans (1989), on the other hand, does provide spectrograms in support of the notion that the redpoll species differ in at least two vocalization types.  He says of the contact calls:

Common [gives a] characteristic “machine gun”-like low and pure chatter, che-che, che-che-che, che-che-che-che-che. Arctic [=Hoary] sounded clearly slower, higher-pitched and less pure, djeet, djeet-djeet, djeet-djeet-djeet.

He illustrates these with spectrograms from two individuals of each species:

Contact calls of two Common Redpolls, from Herremans 1989

Contact calls of two Hoary Redpolls, from Herremans 1989

The two Commons (D and E) resemble each other on the spectrogram, and the two Hoaries (A and B) more or less resemble one another, but there are clear differences between the species, according to this graphic.

Case closed?  No, not quite.

You see, redpolls are cardueline finches, and Paul Mundinger showed way back in the 1970s that in this subfamily, many calls are learned, not innate.  Members of pairs change their call types to match one another, as a bonding mechanism.  Mundinger (1979) paired a captive female Common Redpoll with a male Pine Siskin, and a male Common Redpoll with a female Eurasian Siskin.  In both pairs the two members ended up with nearly identical flight (contact) calls.  In these cases Mundinger wasn’t certain which individuals had changed their calls to match the other’s, but he thought it likely that the male Pine Siskin had switched calls to match his redpoll mate, and that the male redpoll had matched calls with his Eurasian Siskin partner.

Herremans certainly knew of Mundinger’s work.  In fact, he replicated it: he demonstrated that three individual redpolls (a male Common, a female Common, and a female Hoary) all adopted identical “breeding calls” when paired with one another in captivity.  But he didn’t seem to appreciate the implications of this work, which called his earlier claims of species differences into serious question.  Herremans seemed to believe that call matching only happened with “breeding calls” among mated pairs, but Mundinger’s work was with the flight/contact calls, and at least in one species (Pine Siskin), he showed that even two males matched calls when caged together.  (Two females, however, did not.)

So redpoll calls are learned.  Does that mean there can’t be consistent differences between the species?  Not at all.  But if the calls are learned AND flockmates match one another’s calls, we might expect to see flock-specific call dialects.

And that’s exactly what we see when we compare online recordings.  There’s tremendous variation from one Common Redpoll to the next:

 Common Redpoll, Okanagan Co., Washington, 12/22/2011. Click to listen  Common Redpoll, Grand County, Colorado, 1/15/2013. Click to listen  Common Redpoll, Parkland County, Alberta, 4/22/1974. Click to listen  Common Redpoll, Lac La Biche County, Alberta, 6/6/1981. Click to listen

And from one Hoary Redpoll to the next:

Hoary Redpoll, Safety Sound, near Nome, Alaska, 5/14/2013. Click to listen Hoary Redpoll, Kougarok Road, near Nome, Alaska, 6/4/2013. Click to listen Hoary Redpoll, Dalton Highway, Alaska, 6/10/2006. Click to listen

Within flocks, redpolls tend to sound the same.  Between flocks, they tend to sound different.  These differences can be pretty large in geographic scope: I got recordings of three different redpoll flocks last winter from three different Colorado counties, and they are all pretty similar to one another.  But they are quite different from the redpolls in Washington and Alberta.

This is the pattern we would expect to see if flockmates match one another’s calls.  We would expect birds in one flock to converge on a common pattern, while birds in another flock converge on a different common pattern.  If redpolls respond preferentially to birds of a similar call type (like crossbills, their closest relatives within the Carduelinae), that would explain why Herremans’ captive Hoary Redpolls showed little response to the calls of wild Commons flying by.  Maybe it was because they weren’t conspecifics — but maybe it was just because they weren’t flockmates.


There may be consistent differences between the vocalizations of the redpoll species.  Perhaps they use vocal differences to reject mates of the wrong species.  Perhaps we can use differences to tell the species apart.  But we can’t do that on current knowledge.  There simply aren’t enough available recordings to document the limits of variation within species, in order to hash out the differences between species, if they exist.

There’s also the question of whether Hoaries and Commons in the same winter flocks may match each other’s calls.  We don’t know whether they do, but given the willingness of these species to match calls of even Pine and Eurasian Siskins in captivity, it’s at least possible.

So it’s premature, I think, to try identifying redpolls by voice.  But that’s no excuse to ignore it.  We need to learn more about the calls of these birds, and for that we need more recordings.  The next time the redpolls come south, I hope they meet a barrage of microphones.

Empid Mystery Solved!

Empid Mystery Solved!

Remember the mysterious two-part call of the unidentified Empid?  Nacho Areta left a comment this morning on my last post about the sound, pointing out a third example of the mystery sound in the Macaulay Library — this one attributed to Least Flycatcher.  That’s right, the three different recordings of this sound have been identified as three different species — Yellow-bellied, Alder, and Least!

“Whit-beert” call, labeled as Yellow-bellied Flycatcher.
Woodhull Lake, New York, 5/30/1998. ML catalog #106901
“Whit-beert” call, labeled as Alder Flycatcher.
Upton, Maine, 6/2/1962. ML catalog #7546
“Whit-beert” call, labeled as Least Flycatcher. Brown Tract Lake, New York, 5/28/1999. ML catalog #100852

Nacho asked why the calling bird in all three cases couldn’t be Least Flycatcher.  As I went back to review my notes on that species, I realized he’d solved the mystery.  I’d already picked out several examples of what I had tentatively named the “pweet series” of Least Flycatcher, and ML 100852 was one of them — I had simply failed to recognize the resemblance to the mystery call.  Once I went looking for it in the Least Flycatcher collection, I found more examples:

Least Flycatcher. Brown Tract Ponds, New York, 5/30/1999. ML catalog #100876 Least Flycatcher. Bruce Peninsula, Ontario, 5/21/1993. ML catalog #63954

And that last recording is the true Rosetta Stone, because not only does it feature the mystery call, but also many nice unambiguous renditions of the species’ song: the snappy, repeated “che-BEK” that clinches the ID of Least Flycatcher.

Nacho points out that this vocalization type doesn’t have to be a 2- or 3-noted pattern; it’s frequently a longer series.  That, plus the somewhat variable nature of the initial “whit” call, partially explains why these Least Flycatcher recordings were able to sneak right past me, even as I looked directly at the spectrograms.  The entire affair highlights the tremendous usefulness of collaboration in solving mysteries like these!  Thanks to Nacho Areta and everybody else who helped out with comments.