Presented here are data on the known wingbeat rate for Ivory-billed and Pileated Woodpeckers, including take-off, sustained flight, and escape flight.

Summary

1. • The expected wingbeat frequency for Ivory-billed Woodpecker is slower than Pileated based on existing data for woodpeckers in level, non-maneuvering flight (see graph and caveats based on wing morphology under point 3 in “Would level-flight wingbeat rate differ?” below).
   

1. • Wingbeat style (wingtip elevation and depression) as well as flapping rate shows wide variation among individuals depending on conditions and motivating stimulus. Available evidence indicates that Pileated cannot be eliminated as the bird in the Luneau video because it matches or is within the expected range for the species both in flight style and wingbeat frequency.
   

1. • Woodpeckers have different and variable “gaits.” Comparing data for Pileated Woodpeckers exhibiting similar gaits to the bird in the Luneau video, including a sound recording of wingbeats ascribed to an Ivory-billed Woodpecker apparently making a short flight, I find no reason to exclude Pileated.
   

1. i.   take-off wingbeat rate of presumed Ivory-billed Woodpecker  = 8.40 Hz
   

2. ii.  take-off wingbeat rate of Pileated Woodpecker measured the same way = 8.34 Hz
   

3. iii. escape flight for Luneau bird = 8.6 Hz first 5 wingbeat cycles, slowing thereafter
   

4. iv. escape flight for released Pileated = 8.8 Hz first 5 cycles, 8.2 Hz over 9 cycles
   

What is the expected wingbeat rate for Ivory-billed Woodpecker?

Answer: slower than Pileated Woodpecker (but we really don’t know).

Is the take-off wingbeat rate diagnostic?

Answer: probably not.

The only known sample attributed to the wingbeats of Ivory-billed Woodpecker is of a bird apparently making a short flight. I compared the flights of Pileated Woodpeckers making similar short flights to that of the Ivory-billed recording. One sample comparison is presented here.

Ivory-billed Woodpecker

The only sample for this species is the sound of wingbeats recorded by Arthur Allen and Peter Paul Kellogg in 1935 at an Ivory-billed nest site. The recording is not accompanied by images of the bird that is heard flapping its wings. Compared to the entire preserved recordings of these Ivory-billed Woodpeckers, the birds in this segment are comparatively quite agitated. The recorded sounds most closely match a bird making a short flight from one perch to another a short distance away. Only eight wingbeats and sharp thud, presumably the bird landing, can be heard. If one assumes the recorded wingbeats are those of an Ivory-billed, which seems probable, then the sounds can be plotted on a spectrogram and measured. Researchers at the Cornell Lab of Ornithology have done this. I repeated and verified their measurements using Raven sound analysis software.

Pileated Woodpecker

I recorded the wingbeats of a Pileated Woodpecker (subspecies D. p. abeiticola) taking flight and flapping for the same number of beats as the presumed Ivory-billed on the 1935 recording. I extracted audio from the videotape of this flight and plotted the 8 wingbeats on a spectrogram. The sound is not as intense as the Ivory-billed recording because I used an omni-directional microphone on the video camera, not a parabola as used by Kellogg. Measurements of intervals were made with Raven. The spectrogram is presented here.

The measured wingbeat rate for the Pileated Woodpecker here, Dryocopus pileatus abeiticola, is for a subspecies that averages larger (slightly greater average mass) than the subspecies found in Arkansas, D. p. pileatus. Existing empirical evidence shows a strong negative correlation between wingbeat rate and mass (Tobalske 1996). This means that heavier birds have slower wingbeat rates, and less heavy birds have faster wingbeat rates (see graph illustrating wingbeat versus mass, above). It is plausible that the slightly smaller race of Pileated Woodpecker in Arkansas, would be expected to have a faster wingbeat rate than my northern birds. Also note that the interval for each wingstroke is nearly identical for each species over the short duration of these flights.

Conclusion

Take-off wingbeat rates for Pileated and Ivory-billed Woodpeckers are similar. Moreover, as demonstrated below, the wingbeat rate of the bird in the Luneau video is matched by Pileated Woodpecker in escape flight, not just take-off. Therefore, existing evidence does not support the assertion that the bird in the Luneau video is more likely to have been an Ivory-billed Woodpecker based on wingbeat frequency.

Would level-flight wingbeat rate differ?

Answer: possibly, with Ivory-billed having on average a slower wingbeat frequency than Pileated. Importantly, the range of variation in wingbeat rates and other variables (see below) make any diagnosis based on this characteristic very tenuous.

Disregarding overwhelming morphological evidence that the bird in the Luneau video cannot be an Ivory-billed Woodpecker (see Identification), proponents cite wingbeat frequency as a possible feature that eliminates Pileated Woodpecker. Some have said, “show me a video of a Pileated that flies at 8.6 beats per second for a full four seconds." An obvious response to this demand is to ask for any video that shows an Ivory-billed flying at 8.6 beats per second for a full four seconds. None exists, and the Luneau video does not constitute independent evidence for the wingbeat frequency of an Ivory-billed Woodpecker. Since the only measured rate for Ivory-billed Woodpecker (take-off, above) is matched by Pileated Woodpecker, this line of reasoning can be rejected. Are the proponents trying to say that if no video exists of a Pileated making a similar flight, then the bird in the Luneau video must therefore be an Ivory-billed Woodpecker? If so, then this unsupported assertion contradicts available evidence. Moreover, based on current knowledge, the expected wingbeat rate for Ivory-billed would be slower than that for Pileated Woodpecker. Here are several reasons why:

1. •   (1) The only basis for the wingbeat rate of Ivory-bill Woodpecker is a recording of 8 “wingbeats” produced by an unseen bird. The average wingbeat rate for that bird's seven complete wingbeat cycles is 8.4 Hz, not 8.6 Hz, the frequency observed for the Luneau bird. As shown here and in a paper by Collinson (2007, BMC Biol. 2007, 5: 8), this frequency is within the range observed for Pileated Woodpecker.
   

1. •   (2) Available data show that birds flap faster on take-off and then slow down (Tobalske, Nature 2003, 421: 363-366; Collinson, BMC Biology 2007, 5: 8). No data exist to suggest that Ivory-billed Woodpeckers flew faster subsequent to take-off once steady flight was achieved. Therefore, one would expect the sustained wingbeat frequency to be slower than 8.4 Hz and certainly slower than the 8.6 Hz measured for the bird in the Luneau video.
   

1. •   (3) Heavier woodpeckers flap more slowly (Tobalske 1996, Auk 113: 151-177), and the Ivory-bill was the heavier bird. The expected mean wingbeat frequency for steady flight (non-maneuvering flight) of Ivory-billed would could be 4.7 Hz (using a mass of 454g for Ivory-billed in the regression presented by Tobalske 1996). This is much slower than 8.6 Hz and slower than the mean of 5.2 for Pileated Woodpecker. Many variables contribute to wingbeat rate, including aspect ratio (length to width of wing) and length of wing-arm. Birds with higher aspect ratio tend to flap at a faster rate. Ivory-billed Woodpecker has a higher aspect ratio wing than Pileated Woodpecker (see generic description for Pileated under “Phloeotomus” [page 153] and Campephilus [paged 164] in Ridgway 1914, who noted the relatively shorter secondaries of C. principalis, C. bairdii, and C. imperialis, which taxa constituted the whole of the genus as then recognized and which form a distinct, monophyletic group, see Fleischer et al. 2006). Ivory-billed has a longer wing, however, which would contribute to a slower wingbeat rate. The regression analysis found by Tobalske incorporates many wing shapes, including woodpeckers with higher aspect ratio wing shapes, e.g. a sapsucker. How these and other variables might affect the mean wingbeat rate for Ivory-billed Woodpecker is unknown. It appears highly unlikely that higher aspect ratio would overcome the stronger effect of mass and length of wing arm, both of which contribute to a slower wingbeat rate. It should be noted that the values presented are merely predicted values based on the regression equation. Since the values lie outside the span of observed data, they should be interpreted with caution.
   

1. •   (4) Available data do not support the assertion that wingbeat frequency could be diagnostic: individual variation is so great that diagnosis to species is not possible within the range exhibited by Pileated Woodpecker. Statistical analysis shows that one cannot reject a frequency of 8.6 Hz for Pileated Woodpecker based on available data for birds in steady, non-maneuvering flight alone (unpublished analysis by C. Elphick based on original data provided by B. Tobalske).
   

1. •   (5) Although some claim that the bird in the Luneau video sustains a wingbeat rate of 8.6 Hz for over four seconds, the measurable duration is barely 2 seconds in that video. Beyond that point, the image is so blurred that one cannot accurately measure wingbeats. Moreover, my own analysis indicates that the bird in the Luneau video begins slowing within this first 2 seconds (measuring angle of wing in successive wingbeats).
   

Is escape flight wingbeat rate different?

Answer: no.

In early fall 2006, I videotaped the release of three different Pileated Woodpeckers. These were released by hand at the same height as the bird videotaped in Arkansas. Although many differences exist in the parameters at launch between a bird launching itself from low over water and those released by hand over a terrain that drops away, I found that Pileated Woodpecker can match the rate exhibited in the measurable portion of the Luneau video. This is true for both the individual wingbeat rate and averaged wingbeat rates (time averaged over several wingbeats rather than measured for a single cycle).