In my last post, I discussed differences in the alarm calls of Gray Hawk (for the purposes of this post, Buteo plagiatus) and Gray-lined Hawk (here, Buteo nitidus — see the last post for taxonomic rationale). Over the past few days, I have gone through the Gray Hawk sound collection at the Macaulay Library, which has given me some new information.
In my last post I said the alarm calls of Gray and Gray-lined Hawks tend to be “fairly stereotyped within populations.” In light of what I’ve learned from the larger sample size, I think it’s worth retracting that statement in favor of something like, “the alarm calls of Gray and Gray-lined Hawks are variable within populations, and even within individuals, but are nonetheless consistently separable on several characters.”
Of the characters I mentioned in my last post, the most consistent seems to be energy spectrum; those differences seem consistent across the entire sample size. Differences in frequency and pattern of inflection are also pretty reliable, although there are anomalous birds here and there. The differences in duration are tricky, though, since several Gray-lined Hawk vocalizations from Venezuela clock in at around 1.75 sec in length. It’s clear there is a lot of overlap in this character.
Here are some other ways in which the alarm calls vary:
The noise content of alarm calls in both Gray and Gray-lined Hawk is highly variable. A clear tone quality (that is, without noise) is the norm in both taxa, but noisy calls are not infrequent, and some birds in both taxa may give noisy and clear calls on the same recording.
Within the context of the constant differences in energy spectrum, phase shifts (period doubling, period tripling, etc.) are common in both taxa, as they seem to be in many species of raptors.
I hope to publish more on Gray Hawks (and other bird sounds) soon, but as I’m in the middle of an unplanned trip, I may not get to it until next week. Sorry!
Gray Hawk, B. plagiatus, Belize, 2005. Photo: The Lilac-breasted Roller (Creative Commons Attribution 2.0).
The Gray Hawk, whose scientific name recently changed from Asturina nitida to Buteo nitidus (AOU 2006), was historically considered two species:
Gray Hawk (Asturina plagiata or Buteo plagiatus), comprising a single subspecies found from the southern United States south to northwestern Costa Rica;
Gray-lined Hawk (Asturina nitida or Buteo nitidus), including races costaricensis, nitida[-us] and pallida[-us], ranging from southwestern Costa Rica south through much of South America to northern Argentina.
In recent years the AOU has recognized only a single species, with a range from Arizona all the way to Argentina, but evidence is mounting that it should in fact recognize two. The BNA account (Bibles et al. 2002, subscription required) mentions many well-established differences between the two groups in both plumage and measurements, which led B. A. Millsap to propose a split in his 1986 master’s thesis (abstract). Then, in 2003, Riesing et al. published a molecular phylogeny of the genus Buteo which not only found that Asturina was embedded within Buteo (resulting in the aforementioned change of genus) but also found evidence for a species split:
Substantial variability was detected within B. nitidus. The subspecies B. n. plagiatus is 9% apart from B. n. nitidus and B. jamaicensis costaricensis, respectively. Thus, the earlier proposed species status of plagiatus (Millsap, 1989) is supported by our data.
The American Ornithologists’ Union Checklist Committee, however, wasn’t convinced:
Riesing et al. (2003) suggested that the groups should be recognized as distinct species, but did not provide supporting data.
I’m no molecular biologist, so I can’t comment in depth on the quality of the data or the analysis, but I can say that Riesling et al. had a pretty small sample size: only a single sample each of the Gray Hawk subspecies nitidus, costaricensis, and plagiatus. And although they looked at two genetic markers overall, only one marker seems to have been sampled from all three taxa. In that one genetic marker, however, the samples of “Gray” and “Gray-lined” Hawks were as different from each other as the northern “Gray” Hawk is from Red-tailed Hawk, which is a substantial difference indeed.
Here is added evidence for the split: the plagiatus group and the nitidus group differ substantially in vocalizations, a fact that seems to have gone unreported in the literature so far. (Wikipedia says the vocalizations of both groups are identical, but does not cite this information.) In this post we’ll start simple:
One-note (alarm) call
The most commonly heard vocalization of both species is the single-noted call, which apparently indicates alarm, at least in the plagiatus group. The single-noted call appears to be fairly stereotyped within populations, but strikingly different between them:
Alarm call of Gray Hawk (Buteo plagiatus), Guadalupe Canyon, AZ, 5/16/2009 (14-20).Alarm call of Gray-lined Hawk (Buteo nitidus), Rio Caura south of Maripa, Bolivar, Venezuela, 1/4/2007 (03-07).
The following differences can be used to separate the two calls consistently:
Duration. The call of the Gray Hawk lasts 1.5 – 2.0 sec, while the call of the Gray-lined Hawk is typically 0.5 – 1.0 sec long.
Frequency. The call of the Gray-lined Hawk is significantly higher-pitched, with a maximum frequency of ca. 3.5 kHz, versus a maximum of less than 2.5 kHz for Gray Hawk.
Energy spectrum. In the call of the Gray-lined Hawk, the energy is overwhelmingly concentrated in the fundamental, giving the call a clearer, more piercing tone quality, while in the call of the Gray Hawk, it’s the first harmonic that is strongest, with comparatively more energy in the second harmonic as well, relative to the fundamental. This gives the call of the Gray Hawk a distinctly more nasal tone quality.
Pattern of inflection. The shape of the two calls on the spectrogram is consistently different. The Gray Hawk’s call is always a smooth overslur with a very early, very brief peak, with the final 75% of the call comprised of a long, nearly monotone “tail.” The Gray-lined Hawk, meanwhile, always shows a much more evenly overslurred shape, with a “flat-topped” look and a shorter and more sharply downslurred “tail.” In addition, the Gray-lined Hawk almost always has a “break” in its voice, a point in the terminal downslur at which the pitch drops almost instantaneously. In the above example this occurs almost exactly at the 1.5 sec mark.
On this page you will find comparative spectrograms from Xeno-Canto, with (as of this writing) 13 examples of Gray-lined Hawk’s alarm call and 2 examples of Gray Hawk’s. Xeno-Canto’s one recording from Panama, apparently the only representative of the Gray-lined subspecies costaricensis, seems odd, so a little more digging is in order there. Other than that, however, the distinctions I noted above hold in all cases, and I don’t note any systematic differences between subspecies nitidus and pallidus.
I’d like to revisit this issue in a future post, to investigate possible differences in the series call and to bring in results from the Macaulay Library‘s website, which appears to be under the weather at the moment. Stay tuned.
I’ve published Lesson Four on trills and beats. Lessons five and six may be a while in coming, since they are slated to deal with nasality and polyphony, which is where things start to get complicated. Stay tuned!
In addition to the comments on my recent pygmy-owl post, I got five private emails, all of which also implicated Merriam’s Chipmunk as the likely source of the pygmy-owl-like sound. I emailed Doug Von Gausig, and he was very amenable to the possibility that he might have recorded a chipmunk instead of a pygmy-owl. I think this is likely the case.
The best evidence came from a 1976 paper by Leonard Brand in the journal Animal Behaviour: “The vocal repertoire of chipmunks (Genus Eutamias) in California” (email me for a PDF copy). In addition to the “chip” calls that are the most common form of alarm note in all ten chipmunk species in California, Brand mentions a “chuck” call that matches our mystery sound:
Chucks were lower pitched than chips, with their fundamental frequency between 0.5 and 2.0 kHz. Their lowest frequencies were at the beginning and end, with higher frequencies in the middle of the syllable. Eutamias townsendii chucks had from zero to five harmonics at approximately 2 kHz intervals. Each chuck was from 0.03 to 0.05 sec long. Chucks were given by all species, and they all had a form similar to E. townsendii [emphasis mine]. Chucks were usually given…in a steady series, from 50 to 178 per min.
The description is a pretty good match for the pygmy-owl-like sound, but I think the real clincher is Brand’s spectrogram of the “chuck” call:
"Chuck" call of Townsend's Chipmunk (Tamias townsendii), adapted from Brand (1976), Fig. 5.
Probable "chuck" call of Merriam's Chipmunk (T. merriami, Chews Ridge, Monterey County, CA, 1/12/2009. Recording by Andrew Spencer.
I fiddled with the axes on the spectrogram of Andrew’s recording to make the two as comparable as possible, and although there are slight differences, the similarities are striking, right down to the faint third voice in between the fundamental and the first harmonic.
Brand was using a Kay Electric Company spectrograph machine on the wide-band setting, which explains the much thicker lines on his spectrogram. Although I’m not a fan of the wide-band setting, I do have to applaud Brand for resisting the temptation to trace his spectrograms, as was the rage in the 1970s. Thank goodness we have computers to do the dirty work for us these days!
Overall, I see no reason not to identify Andrew’s recording as the “chuck” call of a chipmunk, which would make it a Merriam’s Chipmunk (Tamias merriami), the only species in Monterey County. Here’s another link to the sound:
I still marvel at the uncanny resemblance to an owl!
By popular demand, here’s a natural extension of my first post: an index to the Red Crossbill call types on some of the cuts in the Macaulay Library collection. First, a couple of introductory notes and caveats:
Typing Red Crossbills from recordings is not an exact science. It’s not difficult if the bird is giving actual flight calls, but if the bird is singing, things can get confusing in a hurry, because the songs of Red Crossbills frequently contain strings of repeated sounds that resemble flight calls, but are really just notes in the song. These song notes might be variations on the bird’s own flight call type, or they might somewhat resemble another flight call type, or they might just be something else entirely. The point is that a solid identification to type requires a good string of call notes outside the context of song — and not all recordings provide such.
In addition to song and flight calls, crossbills also give other calls, most notably the excitement calls (or “toops”) and the juvenile begging calls (or “chittoos”). The excitement calls do vary from type to type. It is not known whether the begging calls do too. To avoid confusion, I’ve limited the examples in this post to cuts of flight calls only.
All identifications to type have been corroborated by Matt Young at Cornell. Thanks for your help, Matt!
Groth’s (1993) monograph identified eight Red Crossbill types in North America. Benkman (1999) identified a ninth type endemic to the South Hills of Idaho. Ken Irwin (unpubl.) has proposed that there is a tenth type, which was actually recorded by Groth, but lumped with Type 4. Ironically, my Colorado Birds article mentions this call type as a variant of Type 4 (the variant without the initial downslur). Right now, expert opinions vary somewhat on whether this variant is actually separate from Type 4, but whatever it is, it seems to be relatively common and widespread, or at least widely wandering.
I hope this post is helpful to those who want to try to sort these types out for themselves. In conjunction with the Red Crossbill cuts on Xeno-Canto, the Macaulay cuts should get you off to a good start!
On 12 January 2009, my friend Andrew Spencer recorded an unseen creature on Chews Ridge in Monterey County, California:
Chews Ridge, Monterey County, CA, 1/12/2009. Recording by Andrew Spencer
To my ear this sounds like a Northern Pygmy-Owl (sensu stricto: Glaucidium gnomacalifornicum), and the spectrogram shows many of the characteristics of that species. The note shape is pretty classic, with a sharp initial upslur and terminal downslur, and an overall barely downslurred trend to the rest of the note. Almost all of the energy is concentrated in the fundamental, but a couple of faint harmonics are visible, which is standard for pygmy-owls.
But a few things about this recording are strange:
The notes are quite short, about half the length of the average Northern Pygmy-Owl note;
The rate of the series is quite fast (on average about 1 note every 0.66 sec), which is more than twice as fast as you’d expect from pygmy-owls from California;
The pitch of the notes is just a little high, about 1.5 kHz, while most Northern Pygmy-Owl notes fall just above 1.0 kHz;
Oddly, Andrew reports that the sound was coming not from a tree or bush, but from somewhere on the ground. It’s unlikely that Andrew misjudged the origin of the call: he was aiming at it with a parabolic dish, and if the source of the sound hadn’t been close to the focus of the parabola, most likely the harmonic would be absent from the spectrogram and the echo would be louder in relation to the original sound.
Because the sound came from the ground, Andrew surmised that he might have recorded some kind of mammal. But what kind of mammal could sound so remarkably similar to a pygmy-owl? If there is a mammal in California that sounds like this, then I want to know about it.
I did manage to find a recording that matched Andrew’s: Doug Von Gausig made this recording in almost the same place: along the Carmel Valley Road just east of Carmel, California, on 24 March 1999. In an email, Doug told me his pygmy-owl was unseen, but positioned high within the trees in a dense gallery forest.
The bird on Doug’s recording is even higher-pitched and faster than the one on Andrew’s. Together, these two recordings represent quite a departure from what I’ve come to think of as the “typical” song of californicum. Xeno-Canto, the Macaulay Library and the Borror Lab, between them, have nine recordings of the primary song of Northern Pygmy-Owl, from California, Oregon, Montana, Colorado and Utah. None of the birds in these recordings quite match the Spencer/Von Gausig birds in note length or pitch, and most strikingly, none of them come close in rate: the Borror bird from Utah and Jason Beason’s Colorado bird average about one note every 1.5 sec, while the rest, including the Montana bird, average a note every 2.0 sec or more. At the moment, I’m still a little skeptical that “interior” Northern Pygmy-Owls sing significantly faster than “coastal” Northern Pygmy-Owls on average, but I have long agreed that “Mountain” Pygmy-Owls (Glaucidium gnoma gnoma), which breed from southern Arizona south through Mexico, sing very differently from californicum and probably deserve species status. Here’s their standard song:
"Mountain" Pygmy-Owl (Glaucidium gnoma gnoma), Big Bend National Park, TX, 3/28/2008.
In addition to the typical song with the irregular rhythm, “Mountain” Pygmy-Owls apparently sometimes give a faster, stricter song that is very similar to the song of the Northern Saw-Whet Owl (so similar, in fact, that I suspect the two may represent a vastly underrated ID problem). See the Xeno-Canto forum discussions (here and here) of Allen Chartier’s seven cuts of fast Mountain Pygmy-Owl song from Oaxaca. I was particularly struck by Rich Hoyer’s understanding of a stepwise slowing trend in pygmy-owl song rates as you move counterclockwise from Mexico to Colorado to California to Baja California Sur. Up to a point, that jives with the information I have, but what to make of the two birds from Monterey County that sing at a rate of about 80 notes/minute? Are they representatives of an anomalous local dialect? Or of a different type of vocalization than the primary song? Are they not even pygmy-owls at all?
Last weekend in Minnesota I observed a juvenile Brown-headed Cowbird (Molothrus ater) being fed by a much smaller adult Chipping Sparrow (Spizella passerina). I’ve known for years that cowbirds are brood parasites, which means they lay their eggs in other birds’ nests and leave the job of child-rearing to unwitting foster parents. But this was the first time I’d actually seen it, and heard the cowbird’s begging calls:
Begging calls of a fledgling Brown-headed Cowbird, Murphy-Hanrehan Park Reserve, Savage, MN, 7/6/2009 (19-45).
Full disclosure: I didn’t actually record the Chipping Sparrow’s foster child; the above recording was made the following day, when I chanced upon another fledgling cowbird, this one being fed in underbrush by a parent I never identified.
Hearing two begging cowbirds in two days got me thinking. How important are begging calls to parent birds? What if the begging of the cowbird sounds nothing like the begging of its foster parents’ biological chicks? Is it less likely to be fed?
Because brood parasitism is such a fascinating phenomenon, one with significant implications for both evolutionary theory and for endangered species management, it has been extensively studied. Some brood parasites apparently mimic their host species in both sight and sound, presumably so that the foster parents will be more likely to accept them, but according to Lorenzana & Sealy (1996), Brown-headed Cowbirds do not do this. Over 140 bird species have been known to successfully raise Brown-headed Cowbird chicks, but those chicks always sound like typical cowbirds.
As the Fledgling Project makes clear, baby birds of different species can sound completely different. Therefore, it would seem that parents who respond preferentially to chicks of their own species would have an evolutionary advantage over those who respond willingly to cowbird chicks. It’s been shown that the begging of cowbird chicks tends to be both louder and more frequent than the begging of their non-cowbird nestmates, and it’s been suggested that having a cowbird chick in the nest makes the nestmates change their calls in response, but as far as I can tell, researchers haven’t yet directly tackled the question of differential parental response to Brown-headed Cowbird chicks based on their vocalizations.
Also, I neglected to mention it before, but I’ve been working on updating the Bird Sound Glossary as well. As I expand the spectrogram pages, I’ll add more “example” links to the glossary, and I’ll expand the number of terms as we go.
Ever wondered what a Dodo sounded like? Or a Great Auk? Or a Kaua’i ‘O’o?
On a whim, I went looking for sounds of extinct birds on the internet, and I found a lot more than I bargained for — in more ways than one. I managed to turn up some of the rarest, most remarkable, saddest and most haunting recordings I’ve ever heard…and also some of the looniest. I’ll save the loony for last; let’s start with the poignant.
Dusky Seaside Sparrow
Handsome and doomed: the last Dusky Seaside Sparrow died in 1987. USFS photo (public domain).
On its Florida Bird Sounds page, the FMNH posts links to .wav files of the songs of many Florida Birds, including the Dusky Seaside Sparrow (Ammodramus maritimus nigrescens), of which the last individual died in captivity in 1987 at Walt Disney World. You might argue that they weren’t amazing songsters, but I rather like their style. In addition to the FMNH site, the Macaulay Library has an extensive collection of Dusky Seaside Sparrow recordings.
Ivory-billed Woodpecker
Macaulay also has a number of other extinct bird sounds. One of its crown jewels is, of course, Arthur Allen’s famous 1935 recording of Ivory-billed Woodpeckers (Campephilus principalis) in Louisiana. Interestingly, a search in the Macaulay catalog for “Ivory-billed Woodpecker” also turns up John Dennis’s 1968 recording from Texas, a cut that has been the subject of much debate — see the “Notes” section of the linked page for a hint of the history of the discussion.
Bachman’s Warbler
This was one of my most exciting finds. Both the Macaulay Library and the Borror Lab have copies of the two known recordings of Bachman’s Warbler (Vermivora bachmanii), from 1954 and 1959, some of the species’ final years. One of the reasons I find the song of this species so fascinating is that it contains more than an echo of the more familiar sounds of its surviving relatives, the Blue-winged and Golden-winged Warblers. In short, in spite of everything, it sounds ordinary — which just makes it the more remarkable.
Hawaiian Birds
It’s not surprising that Hawaiian birds should make up a good percentage of the extinct species for which recordings exist: first, because Hawaii has lost so many species (around 35 since 1800) and, second, because a number of those extinctions have occurred in recent years, during the era of audio recording. Here is a sampling of Macaulay’s treasures:
Po’ouli (Melamprosops phaeosoma), extinct since 2004: one cut online
Hawaiian Crow (Corvus hawaiiensis), extinct in the wild since 2002: five cuts online
There may well be other extinct Hawaiian bird recordings out there, but I didn’t have time to track down any more. Note that the Borror and Macaulay search functions, and likely other internet searches as well, often have difficulty with the apostrophes that are ubiquitous in Hawaiian bird names, so searching the scientific names tends to work much better.
Spix’s Macaw
Xeno-canto has a couple of recordings of the last wild individual Spix’s Macaw (Cyanopsitta spixii) from the year 2000. Like the Hawaiian Crow, this species persists in captivity, so hopefully its absence from the wild is temporary.
One of the most remarkable things about the Huia was the sexual dimorphism of its bill: males (front) had short, straight bills, while females (rear) had very long, decurved bills. Painting by Walter Buller, 1888 (public domain).
Huia
The Huia (Heteralocha acutirostris), endemic to New Zealand, was last seen in the early 20th century, so no recordings of its song exist. However, in the early 1990s, British composer David Hindley attempted to recreate its song using a computer. Hindley’s efforts were based on a pretty solid foundation: in 1954, the New Zealand Broadcasting Corporation had made recordings of an elderly Maori man named Henare Hemana (or Hamana or Haumana; sources differ on his name), who remembered the song of the Huia from his youth and was able to whistle an imitation of it for the recorder. Hindley used that recording, together with recordings of extant New Zealand birds, to try to recreate the Huia’s song as accurately as possible. You can’t listen to Hindley’s work on the web, as far as I know, but you can hear the original recording of Hemana’s imitation here (by playing the embedded video). Hemana’s whistles sound otherworldly, beautiful, and entirely appropriate for a bird as unique and highly prized as the Huia.
Interestingly, after he recreated the Huia’s song, Hindley was commissioned to recreate the sounds of the extinct Dodo (Raphus cucullatus) of the island of Mauritius. Hindley admitted that his Dodo song was “a fantasy,” because almost no information about the bird’s actual call has survived, but he used what he knew about doves (the Dodo’s relatives) and his own imagination to create a sound that included bass-register cooing, “screaming” and “percussive sounds.” Wouldn’t you love to hear that in a Mauritian forest? Unfortunately, like his Huia reconstruction, Hindley’s Dodo sounds don’t seem to be available on the web.
Séance Vocibus Avium
The coo/scream/bang of Hindley’s Dodo provides us with a nice transition into the looniness I promised earlier. Perhaps taking inspiration from Hindley, a German avant-garde musician named Wolfgang Müller in December 2008 published a collection of reconstructed sounds of extinct birds called Séance Vocibus Avium. Packaged with a 40-page book, the CD contains eleven tracks, one per species, each authored by a different avant-garde musician. Unlike Hindley, Müller et al. didn’t have whistled imitations to start with; they worked only from published voice descriptions, some of them very old. The list of species is quite diverse:
As of this writing, all eleven of these tracks can be heard on the internet. They were played on the 22 January 2009 broadcast of The Wire’sAdventures in Modern Music, which can be downloaded or streamed; tracks 1-5 can be heard from 31:30 to 36:45, and tracks 6-11 can be heard between 1:19:15 and 1:24:30 (update: those links are now broken, but you can hear four of the tracks here).
Warning: you may not recognize the other sounds on this broadcast as music. You may get particularly confused at 1:19:15, as the recreated sound of the Mauritius Blue Pigeon begins to crossfade in over the end of the previous track, which is called “Extensity Of Hard Disk Drive” and indeed sounds like something my computer hardware might sing if I were using it as a percussion instrument. Except for the pigeon, all the other tracks are introduced by a man’s voice speaking the common name of the bird in German, much as you would expect on a commercial bird sound CD.
My reaction to these tracks was decidedly mixed. I liked the ‘o’o, the pigeon, the rail and the caracara. The Pink-headed Duck and the Heath Hen sounded to me like people imitating birds. In the cases of the Jamaica Petrel and the Lord Howe Starling, the artists seem more interested in their art than in any real attempt at verosimilitude: the petrel “call” harkens forward to a time when humans have solved the extinction crisis by introducing cybernetic replicas of endangered species into the wild, and the guy who did the starling was just plain frightening.
I have to say, though, that the grand finale, the Great Auk, was well worth the wait. It also sounded like a person imitating a bird, but with masterful abandon, squawking and gurgling in ways I think would have been worthy of the Original Penguin itself. Great fun.
I’d love to hear of more sounds of extinct birds on the web, if anyone finds others.
