Supporting Online Material
1. Materials and Methods.
The 25 April 2004 video (the "Luneau video") was obtained using a Canon GL-2 video camera, recording
digital video–compressed video onto MiniDV videotape with interlaced NTSC (29.97fps/480i) format. The
camera was set to "normal" mode at fully automatic shutter speed and aperture. Skies were overcast. The
continuously running camera was mounted near the center of a slowly moving canoe just above gunwale
level and was focused at approximately 2 m. The lens was at its greatest wide-angle setting, which records
a field equivalent to that of a 39.5 mm lens on a 35 mm camera. The bird was approximately 20 m away as
it started its flight.
Video recordings were captured from the miniDV with Apple Firewire and FinalCut Pro HD software.
Frames were extracted to image sequence via QuickTime and converted to uncompressed TIFF format
(720x480 pixels). Frames were deinterlaced via Mathworks Matlab and Adobe Photoshop, in both cases
creating two daughter fields (different 720x480 images with alternate odd- and even-field scan sets) per
frame, then cropped and resized. Qualitatively identical results were produced from Photoshop (with the
"deinterlace video" filter and bicubic interpolation) and Matlab (with either bicubic or bilinear interpolation
for simultaneous deinterlace and resize). No other modifications, photo retouching, contrast adjustment, or
video-enhancement were performed.
We compared features of the Luneau video with museum specimens of ivory-billed and southern pileated
woodpeckers (subspecies Dryocopus pileatus pileatus), archival photographs and film frames from the
Singer Tract birds taken by A. A. Allen and J. T. Tanner (S1), and a detailed re-enactment with life-sized,
accurately painted models of ivory-billed and pileated woodpeckers, which we conducted at the scene of
the original video.
For the re-enactment (performed 15 March 2005), we constructed, painted, and mounted on long poles
lifelike wooden models of ivory-billed and pileated woodpeckers with outstretched wings. Operators could
flap the wings at a rate of about 1 flap s–1. Models were held at different angles to reveal both upper and
lower wings. Slow shutter speeds (1/8 s, 1/15 s) produced blurred images comparable to those on the
Luneau video. The re-enactment's location, camera-to-subject distance, light conditions (overcast sky near
midday), camera (Canon GL-2), and studio treatment were the same as in the original. Both woodpecker
models show more white in the re-enactment video than is actually present, an artifact of the dominance of
white in video imaging under normal outdoor conditions. As a result, the underwing of the ivory-billed
woodpecker model appears solid white. However, the underwing of the pileated woodpecker model
consistently shows a trailing black edge.
We estimated the wingspan of the bird in the Luneau video from the distances between outermost points of
white on the outstretched wings on consecutive wingbeats 3 to 12. We corrected measured values upward
to account for the black primaries in ivory-billed and pileated woodpeckers (indistinct and unmeasurable at
mid-wingbeat in the video; correction factors 1.35 and 1.31, respectively), and we calibrated them using
the width of the tupelo trunk. At the first fully visible wing beat (third overall) (Fig. S3) the apparent
wingspan on the video is 70 to 71 cm. True wingspan must be larger, because the distance from camera to
flying bird was greater than that to the tupelo trunk used for calibration.
In our acoustic monitoring, we deployed autonomous recording units (ARUs) at 16 sites from 21 March to
13 May 2004 (deployments of 4 to 49 days recorded 2693 hours) and at 86 sites from 18 December 2004
through 1 April 2005 (deployments of 8 to 30 days have recorded 9395 hours to date). Each ARU contains
a programmable microprocessor, 12-bit analog-to-digital converter, omnidirectional microphone,
preamplifier, signal conditional circuitry, and a hard disk. The components are housed in a waterproof
polyvinyl chloride container, attached to a tree trunks 2 to 3 m above ground, and powered by an external
battery pack. The detection radius for ivory-billed woodpecker is unknown, as we lack data on their
acoustic source levels. The detection radii for other woodpecker species are 150 to 200 m.
Between 12 March and 10 May 2004, 30 individuals spent about 2500 party-hours surveying 15 km2 of
swamp forest in the vicinity of the initial sighting. Observers wore camouflaged clothing and made slow,
quiet movements or remained stationary for periods at lake edges and powerline cuts. Between 25 May and
20 December 2004, T. Barksdale spent portions of 165 days (9 to 12 hours per day) observing from a 25 m
elevated boom at the forest edge. Between December 2004 and April 2005, we worked in the Cache-
White-Arkansas river area with 20 full-time observers and an average of 2 additional rotating volunteers.
We devoted 4750 party-hours to searching 145 km2 of forest for foraging signs and roost or nest cavities
by walking global positioning system–guided parallel transects spaced 50 to 55 m apart. Cavities of
suitable size were watched from 85 min prior to sunset until 10 min after sunset. Our searching was
focused in areas of mature forest that contain concentrations of dead and dying trees. On 2–4 March 2005
we conducted a "saturation search" with 23 observers spaced across the 4 km2 of forest in which all the
sightings had occurred. We also used eight video-camera traps, which were placed at possible foraging
sites over various intervals. Broadcast playback of ivory-billed woodpecker vocalizations was used
2. SOM Text.
Reports with disputed evidence. Reported observations and recordings of at least one ivory-billed
woodpeckers in Highlands County, Florida, from 1967 to 1969 ended when the tree containing their cavity
blew down. A white feather reportedly recovered from near the downed cavity was identified by Alexander
Wetmore as an innermost secondary of an ivory-billed woodpecker (S2). The nest cavity and feather are
archived at the Florida Museum of Natural History, Gainesville, Florida. Audio recordings reported to be
from the birds they observed (archived in the Macaulay Library at the Cornell Laboratory of Ornithology)
are actually pileated woodpecker, northern flicker (Colaptes aurata), and red-shouldered hawk (Buteo
lineatus). Calls recorded by John Dennis in the Big Thicket, eastern Texas, in 1968 (S3) are considered by
many experts as not convincingly separable from those of blue jay (Cyanocitta cristata). Two photographs
taken in the Atchafalaya Basin, Louisiana, in 1971 (S4) that show an adult male ivory-billed woodpecker
perched in similar posture on two different tree trunks are suspected by many ornithologists to be a hoax.
11 February 2004 sighting. Field marks noted by G. Sparling were the bird's unusually large size
compared to pileated woodpecker, peculiarly pointed red crest with black anterior edge, long neck, and
extensive white on lower half of folded wings showing slight yellowish tinge along edges "like parchment
27 February 2004 sighting. T. Gallagher and B. Harrison both noted (with naked eye at a distance of 20
m, measured later) that the basal two-thirds of the flight feathers, including all secondaries and several
inner primaries, were snow-white in sharp contrast to jet-black elsewhere on the wings and along a narrow
stripe between the two outstretched wings (Fig. S1). Having fixed their attention on the striking and
diagnostic pattern of white in the wings, neither observer noted any features of the head or upper back.
Other convincing sightings from the Big Woods. Of at least 15 reported visual encounters between 11
February 2004 and 14 February 2005, seven contained sufficient detail for the authors to treat them as
authentic. In addition to the two February 2004 sightings described above, these were: 5 April 2004 (James
M. Fitzpatrick saw overflight along a lake edge; with naked eye at 100 m he noted large size, broad white
trailing edges of wings, and steady "loon-like" flight of otherwise black woodpecker); 10 April 2004
(Melinda LaBranche saw overflight at site of 5 April sighting; through 10-power binoculars at 100 m she
observed broad white trailing edges of wings and narrow red crescent on rear of folded crest); 11 April
2004 (Melanie Driscoll watched a large woodpecker fly across a 50-m gap in the forest where she was
stationed, and through 10-power binoculars at 120 m she saw broad white trailing edge of wings, white line
extending from wings up the long neck, and small flash of red on crest, with head otherwise black); 9 June
2004 (Harrison saw a large woodpecker flush from near the base of a bald-cypress, Taxodium distichum,
about 15 m in front of him, and with naked eye he noted broad white trailing edges to wings, especially
visible as the bird swooped upward to land; 14 February 2005 (Casey Taylor heard double-rap display
drums for 30 min near power-line cut and then saw a large black-and-white woodpecker crossing the cut
while being mobbed by American crows, Corvus brachyrhynchos. Through 8-power binoculars at 80 to
120 m, she noted broad white trailing edges to wings, long neck with white stripe running its full length,
and black head with long bill.)
Audio evidence. Observers heard series of the distinctive "double-knock" display-drums on three
occasions: 9 November 2004, 14 February 2005 (followed by a visual sighting by the same observer), 7
March 2005 (simultaneously heard by two observers). The series ranged from 3 to 18 double-knocks, and
all three series were heard within 7.5 km of the visual sightings, 10 to 90 min before sunset. The 18
double-knocks heard by a highly experienced observer on 9 November 2004 came alternately from two
directions about 90 degrees apart over a period of 13 min, presenting the impression that two birds were
involved. However, the possibility that a single bird was flying back and forth, alternately drumming from
two different positions could not be ruled out.
 Double-knock sounds strikingly similar to Campephilus display-drums were recorded on several of the
ARUs and appear to be concentrated in the vicinity where drumming was heard on 9 November 2004. We
cannot positively associate these recorded signals as belonging to ivory-billed woodpecker, however, and
several seem out of context. Series of nasal calls closely resembling those recorded by A. A. Allen at the
Singer Tract in 1935 were recorded at two places in the White River National Wildlife Refuge, but these
may have been given by blue jays (Cyanocitta cristata, a notorious mimic). Certainly, if blue jays do
mimic ivory-billed woodpeckers in this region, the implication that a model exists bears investigating.
Home range in fragmented forest. In southern Sweden, black woodpeckers (Dryocopus martius) occur in
densities of 1 pair per 6.7 km2 in contiguous forest, but in landscapes fragmented by agriculture they must
range over 24.3 km2 to cover the same area of forest as pairs in contiguous forests. Pairs in the two
landscapes showed no difference in reproductive rate or predation rate (S5). In Denmark, black
woodpeckers regularly commute up to 6.0 km daily between roost and nest sites (S6).
The following people contributed invaluable time in the search effort: Mike Anderson, Kristina Baker,
Evan Barbour, Jim Bednarz, Catherine Berchok, Bob Berry, Dave Bonter, Mike Braun, Ned Brinkley, Ben
Clock, Jaimie Conrad, Bill Cornatzer, Matt Dresser, Melanie Driscoll, Andrew Farnsworth, Martha
Fischer, Jim Fitzpatrick, Sarah Fitzpatrick, Jeff Gerbracht, Jim Goetz, Julie Hart, Robert Henderson, Bill
Hollimon, Marshall Iliff, Clark Jones, Steve Kelling, Mindy LaBranche, Mike Lanzone, Dan Lebbin, Tim
Lenz, Guy and Joan Luneau, Jay McGowan, Kevin McGowan, David Mehlman, Nick Meyer, Harold
Mills, Jason Mobley, Ryan Mollnow, Lauren Morgens, Edwin Morgens, Allan Mueller, Bob Nixon, Sean
O'Brien, Gerard Phillips, Tina Phillips, Robert Potts, John Puschock, Caleb Putnam, Mark Reaves, Dana
Ripper, Michelle Rogne, Gary Rosenberg, John Ruthven, Matt Sarver, Justin Schuetz, Mickey Scilingo,
Utami Setiorini, Roger Slothower, Carter Smith, Greg Spahr, Timothy Spahr, Doug Stotz, Brian Sullivan,
Casey Taylor, Gerrit Vyn, Ben Wardwell, Michael Warriner, Jeff Wells, Megan Whitman, David Willard.
For support coordinating activities among agencies and organizations we acknowledge: Jon Andrew,
Johnnie Biscoe, Martin Blaney, David Brown, Hank and Cathy Browne, Bob Cooper, Phyllis Dague,
Nancy Delamar, Tom Foti, Jay Harrod, Scott Henderson, Richard Hines, Jerry Hogan, Brad Humber,
Chuck Hunter, Larry Mallard, Jim Peterson, Patty Porupski, Kathy Radomski, Catherine Rideout, Bill
Shepherd, John Simpson, Karen Smith, Allison Wells, Dennis Widener, Max Woodfin, Bob Zielinske.
Diane Tessaglia-Hymes assisted with figure preparation. For acoustic survey and analysis we
acknowledge: Tom Calupca, Harold Figueroa, Melissa Fowler, Tom Fowler, Beth Howard, Anne
Klingensmith, Mike Pitzrick, Dimitri Ponirakis, Mike Powers, Eric Spaulding, Chris Tessaglia-Hymes,
Ann Warde. For loans of field equipment we thank Chuck Hancock, Wade Kreie, Mr. and Mrs. E. M.
Sparling, Jr. To the many dozens of family members and friends who graciously put up with travel and
confidentiality requirements during this project, we owe enormous thanks.
4. Supporting Figures.
Fig. S1. Sketches from field notes of T. Gallagher (A) and B. Harrison (B) that depict the pattern of white
and black they noted on the wings and back of the large woodpecker they observed together in the
Cache River National Wildlife Refuge, Arkansas, on 27 February 2004. Sketches were made
independently of one another.
Fig. S2. Frame from video taken by M. David Luneau on 25April 2004 during a daylong search in the
Cache River National Wildlife Refuge. Red rectangle identifies the area of the video magnified in
Fig. S3, which includes the fleeing woodpecker (video-frame 300 in Fig. S3). Distance from camera
to bird is 20 m. Observer in frame is Robert Henderson, who reported seeing a flash of red in the
fleeing bird, which suggests that it was a male.
Fig. S3. Sixty-five consecutive de-interlaced video fields from the Luneau video, representing 1.08 sec
elapsed time. Fields cover a 72x48 pixel area of the original recording. 10x magnification achieved
via bilinear de-interlacing followed by bicubic interpolation to achieve 720x480 pixels per image.
Frame-numbers represent elapsed time (ms), beginning immediately prior to initial wing-opening by
perched woodpecker. Arrows and imbedded comments identify key moments of the video and
features positively identifying the bird as an ivory-billed woodpecker. The first wing beat is
obscured by the tree. Wing-beat 2 is partially visible, and wing-beats 3 to 9 are fully exposed.
Fig. S4. Distance from exposed area of wrist to the tip of the tail at the moment of take-off as revealed in
the Lunneau video (right-most measurement; see Fig. 1), and comparable measurements on standard
museum specimens of pileated woodpeckers (PIWO) and ivory-billed woodpeckers (IBWO) and
two individual ivory-billed woodpeckers filmed in 1935 by Arthur A. Allen. Scale of the filmed
individuals determined via measurements of the original nest-cavity and trunk features. Bars with tic
marks represent ranges of individual variation among museum skins; bars with open circles
represent variation among independent estimates for the filmed individuals. All estimates for the
2004 bird greatly exceed the largest pileated woodpecker we measured.
Fig. S5. Two treatments from the Luneau video (A and B), which show an indistinct object bordered above
and below by black, along with a video frame (C) and digital photograph (D) taken at the same site
during a 15 March 2005 re-enactment of the scene.In the Luneau video, this object remains in view
for 6 s before the field of view pans away. A fleeing woodpecker (Fig. S3) escapes from near the
same spot 21 s later. (A) 10x bicubic zoom of an otherwise unmanipulated, still interlaced video
frame showing the object on the tree. (B) An average of five consecutive video frames in which the
object was lined up manually and collapsed into a single, slightly clearer image. (C) 10x bicubic
zoom of an otherwise unmanipulated, still interlaced video frame from the re-enactment, showing a
life-sized model ivory-billed woodpecker fixed to the same spot on the tree. (D) Digital photograph
of the model exactly as filmed in the re-enactment to produce the frame in (C). We interpret the
object in the Luneau video as an ivory-billed woodpecker.
Fig. S6. In the "Big Woods" region of eastern Arkansas, USA, mature or sub-mature bottomland forest
covers more than 220,000 ha (green-shaded area), and substantial additional land is being reforested.
5. Supporting References and Notes.
S1. Archival photographs, 35-mm movie, and still images are property of David Allen and Cornell
Laboratory of Ornithlology, Cornell University, Ithaca, New York. Nest-cavity specimen of ivorybilled
woodpecker from the Singer Tract is property of the Cornell Laboratory of Ornithology, Cornell
University, Ithaca, New York.
S2. H. N. Agey, G. M. Heinzeman, Fla. Nat. 44, 46 (1971).
S3. J. W. Hardy, Am. Birds 29, 647 (1975).
S4. J. J. Williams, Birding 33, 514 (2001).
S5. M. Tjernberg, K. Johnsson, S. G. Nilsson, Ornis Fenn. 70, 155 (1993).
S6. H. Christensen, Dan. Ornitol. Foren. Tidskr. 96, 187 (2002).