Sonograms or spectrograms are visual renditions of acoustic phenomena. Cornell’s Raven is perhaps the best-known software for producing these renditions; we use Sonic Visualizer, another program, which is simple and available for free, to generate our sonograms. We use WavePad to snip selections, amplify clips up to 18x, and apply modest filtering or noise reduction. These modifications do not alter the harmonic structure revealed in the sonograms.
In a sonogram, the x-axis represents time, and the y-axis represents sonic frequency. Birdsongs often have complex, distinctive harmonic structures. Ivorybills don’t sing and have a limited repertoire of sounds, all of which have a ladder-like appearance on the sonogram.
Each step in the ladder is called a “partial.” The lowest partial is called the fundamental frequency; the next one is called the second partial, and so on. Sonograms can also tell us about harmonic emphasis and reveal the relative strength of each partial. Subtle changes in pitch within a given note are revealed by the shape of the harmonics. A flat line is a monotone, and inflections can rise or fall or both within a single call.
The x-axis enables us to measure the duration of a sound and the intervals between sounds. At this stage, we have focused more closely on harmonic structure than on duration of calls and spacing between or among sounds because the literature makes it clear that the duration of individual Ivory-billed Woodpecker notes can vary considerably. In addition, distance affects the way sound travels.
The harmonic structure of the Ivorybill call is similar to a wide variety of other sounds, both natural and mechanical. This complicates the process of distinguishing suspected Ivorybill vocalizations from potential confusion notes: squirrels, Blue Jays, nuthatches, and tree squeaks are among the most challenging. The challenges exist whether the reviewer is human (using mixed visual and auditory cues) or automated (using a visual pattern recognition algorithm), but we think we have been able to overcome some of them. The challenge is even greater when it comes to the Ivorybill’s “double knock” (equivalent to other woodpeckers’ drumming, see below), since anything from gunshots to raindrops can mimic the sounds’ visual pattern, and there are no undisputed recordings of Ivorybill double knocks. The remotely recorded sounds discussed in this post were found through a mixed visual and auditory review, with promising sounds first identified by sight.
Part 2 of this post will cover some of the confusion sounds.
For more on avian bioacoustics, we recommend Nathan Pieplow’s earbirding site.
As with so much about the Ivory-billed Woodpecker, our knowledge about the sounds it makes is fragmentary and often conflicting.
The typical call of the Ivory-billed Woodpecker is usually transliterated as a “kent.” In 1935, an expedition by the Cornell Laboratory of Ornithology to Louisiana’s Singer Tract (now the Tensas National Wildlife Refuge and the location of North America’s last undisputed Ivorybill sighting) captured the only universally accepted audio recordings of Ivory-billed Woodpecker vocalizations. This is a sonogram showing kent calls from the 1935 recordings.
The 1935 recordings included the typical kent call and a second call that Tanner described as conversational and usually occurring around a nest. A third type of call that has been described in the literature was not recorded. Additional sounds on the Cornell recordings have not been well described. Adding to our confusion around Ivorybill calls is that throughout the Cornell Lab recordings, the proximity of several researchers to the nest, the presence of a parabolic microphone, as well as a massive piece of recording equipment, on a cart, undoubtedly agitated the birds. As a result, it is possible that the only undisputed recordings of Ivorybills are somewhat atypical. This limited record of Ivorybill sounds is available at the Macaulay Library website.
In 2005, the Cornell Lab reanalyzed another set of potential Ivory-billed Woodpecker sounds recorded by John Dennis in Texas in 1968. An earlier analysis had suggested the sounds were more likely Ivorybill than the sometimes similar Blue Jay. The 2005 analysis pointed more strongly toward Ivorybill but could not completely exclude Blue Jay or White-breasted Nuthatch. This recording, too, is archived at the Macaulay Library website as “Ivory-billed Woodpecker.”
Though the late 1960s reports from Texas are contested, there is nevertheless considerable physical evidence supporting the Ivorybill claims, including the Dennis recording, poor photographs, and the George Reynard sighting and recording (described below). Thus, we suspect the Dennis calls were made by an Ivory-billed Woodpecker, and may in fact be more reflective of what the bird actually sounds like under normal conditions.
In the early 2000s, apparent Ivorybill calls were recorded by Cornell Lab in Arkansas and by Auburn University and the University of Windsor in Florida.
Searchers with Project Principalis have captured numerous suggestive sounds in two parts of Louisiana since 2009. Here are some examples from our current study site; the first two clips were captured by our Audio Moth (remote recorder) deployments in 2019 and 2020, and the last two were recorded by Phil Vanbergen in March 2017.
Acoustic analyses of these and other recordings are ongoing. Searchers with Project Principalis have obtained many thousands of hours of acoustic data over the last two seasons. Manual review of a tiny fraction of these recordings has yielded some promising results, but we await a more comprehensive analysis aided by artificial intelligence (AI) and machine learning programs. Unfortunately, the very simple structure of Ivorybill vocalizations, and their inherent rarity, present a major technical challenge, and the AI being used for the analysis continues to be refined. We hope that it will ultimately provide us with additional strong evidence and information about behavior patterns.
Unlike most other woodpeckers, Ivorybills and others in the genus Campephilus do not “drum” (although a few accounts in the literature suggest Ivorybills do). Instead of drumming, they communicate with “double knocks” (sometimes described as “double taps” or “double raps.”) Other than the recent searches, there is only one recording of a claimed Ivorybill double knock. This recording, which was associated with a sighting, was made by ornithologist George Reynard, in late 1960s in Texas.
Reynard’s recording appears, in a looped format, on a Birds of Cuba CD; it is not well-known in North America.
James Tanner, the Cornell Lab researcher who studied Ivorybills in the 1930s and whose opinions had immense influence, did not find Reynard’s double knock recording convincing. He would have based this assessment on 30 year-old memories.
But Reynard was a respected ornithologist who would later participate in the 1986 Short expedition to Cuba, when he was one of several people to report seeing an Ivory-billed Woodpecker. (The Cuban Ivory-billed Woodpecker is probably a distinct species, based on the best available DNA data. While the two are almost identical in appearance and voice, some still consider it a subspecies.) While no physical evidence was obtained in the Short expedition, these sightings were the last commonly accepted Ivorybill records for Cuba and the continental United States.
Here are three examples of what we suspect are Ivorybill double knocks from the Project Principalis study site. Searcher Patricia Johnson captured one in the field; the others were obtained by a remote unit shortly before sunrise.
We have also heard and recorded possible Ivorybill double knocks in response to anthropogenic double knocks (ADKs). Searchers perform ADKs in the field in hopes of generating a response or an attraction. Such events are rare because birds both have to be in earshot and be sufficiently stimulated to react or fly in. In other Campephilus woodpeckers, the double knock functions as a territorial display, so ADKs are an effective attraction tool. Since the available evidence suggests that even breeding Ivorybills do not defend their territories from other Ivorybills, the function of the double knock may be quite different when compared to South and Central American Campephilus woodpeckers. Thus, making inferences from the behavior of congeners is problematic, and ADKs may be less effective at generating a response or attracting a bird.
Stay tuned for Part 2, which will cover potential confusion sounds.