Woodpecker Drilling Behavior
An Endorsement of the Rotational
Philip R. A. May, MD; Joaquin M. Fuster, MD; Jochen Haber, MS; Ada Hirschman
\s=b\ High-speed cinematograph films of a
drilling woodpecker were examined by
direct visual inspection and by a micro-
densitometer and computer-imaging
technique. These showed (1) that the drill-
ing trajectory is essentially linear, with
very little, if any, rotation of the head; (2)
that there is minimal movement after
impact; (3) that the impact velocity is of
the order of 600 to 700 cm/s; and (4) that
the impact deceleration is of the order of
1,000 g.
Dynamic and morphologic findings in
the woodpecker may be highly relevant to
the prevention of concussion and brain
injury in man. Taken in the context of
modern packaging technology and other
animal and mathematical modeling re-
search, they suggest that brain injury
preventive systems could be greatly
improved over those now in common
use.
(Arch Neurol 36:370-373, 1979)
on the morphology of
observations
"
the woodpecker indicate that the
bird's ability to pound its head
repeatedly on a tree without appar¬
ently incurring concussion or head
injury may be related to firm packing
of the brain, powerful head-neck
muscles, and a narrow subarachnoid
Accepted for publication July 17, 1978.
From the Health Services Research and Devel-
opment Laboratory, Veterans Administration
Hospital, Brentwood, Calif (Dr May, Mr Haber,
and Ms Hirschman), and the Neuropsychiatric
Institute, University of California at Los
Angeles School of Medicine (Drs May and Fuster
and Ms Hirschman).
Reprint requests to Neuropsychiatric Insti-
tute, University of California at Los Angeles
School of Medicine, Los Angeles, CA 90024 (Dr
May).
Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 05/31/2015
Theory of Impact Brain Injury
space with relatively little CSF.1-2 Yet,
although this provides some light as to
how the brain might be shielded from
translational coup and contrecoup, it
does not entirely explain how it might
be protected from impact rotational
forces or from shaking injury during
its rapid to-and-fro movements. Nor
can morphology alone provide an
adequate description of the forces and
movements involved.
This article focuses on dynamic
aspects of the woodpecker's activities
that are important inasmuch as exper¬
imental evidence has not, as yet,
resolved the controversy as to whether
diffuse impact brain damage in
primates is related to translational
linear forces—and the accompanying
positive and negative pressure
waves—or to rotational (angular) mo¬
tion that may produce shearing forces
within the neuroaxis.38
The woodpecker's small size un¬
doubtedly gives it an advantage over
man when it comes to head-pounding
on a tree: Simple theory indicates that
the ratio of brain weight to brain
surface area is low, so that a given
impact deceleration force will be
spread over a relatively greater
out
area. theory, also, small animals
In
have higher resonance frequencies
and should tolerate higher long-dura¬
tion impact loads but smaller impulse
loads.5
Ommaya et al" cited an unpublished
piece of reasoning by Holbourn that
the rotational acceleration required to
produce injury in brains with similar
properties and shapes is inversely
proportional to the two-thirds power
of their mass. Woodpeckers with
brains ranging in weight from 1.25 to
3.95 g should, in theory, therefore, be
50 to 100 times less susceptible to
concussion than man. It seems unlike¬
ly, however, that relative size alone
suffices to explain the bird's apparent
immunity to concussion and brain
injury. Small animals can be knocked
out by impacts that seem to be of the
same order as the blows of the wood¬
pecker, and birds are commonly
observed to tumble motionless (or
twitching a little) and apparently
unconscious after flying head first
into a window, yet they pick them¬
selves up and fly away later (so
demonstrating that their fall was not
the result of a broken neck).
Simple theory based on brain mor¬
phology alone may be misleading. The
Holbourn equation contains many
assumptions: It was developed for
forces of relatively long duration
(more than 6.5 ms) and does not apply
to short-impulse loading; it assumes
that the crucial variable is brain mass
and that all brains have similar prop¬
erties, shapes, and CSF flotation
effect. Further, the woodpecker's
brain is not isolated from its body and
moving independently; the energy
involved is not just a simple function
of the weight of the brain. Our films
(and earlier observations by Spring10)
show that woodpeckers that pound
vigorously throw the full weight of
their body into the action. The impact
is taken by the head, but the energy
involved is a function of the weight of
the body above the femoral neck rath¬
er than the weight of the brain alone.
Furthermore, next to nothing is
known about the cumulative effects of
repeated blows and vibrations. Simple
intuition and clinical observations of
boxers, together with reports on shak¬
ing injury in children,1113 suggest that
 Direction

_y_
„..· ·"
K_.»
I '
...
" . ,,* .ill
'
.
-c ; > >,
¿ '··

C !·'" >', C· -C->~—

Fig 1 .—Woodpecker drilling, plotted frame by frame. indicates strike trajectory of tip of beak; V, vertex; C, head.

the woodpecker's rapid and repeti¬

Table 1 —Drilling Impact: Hand Plots
tious drumming and drilling expose
its brain to particular hazard. Maximum Velocity, Stopping Time, Deceleration,
All in all, it seemed worthwhile to cm/s ms 9
investigate further whether the wood¬ Sequence 1
Inspection method 1,060 <0.5 :2,170
pecker finds it advisable to protect Outline method 748 <1.0 763
itself against rotation injury—or to
Sequence 2
put it more scientifically, whether Inspection method 618 £0.5 >1,260
natural selection has favored the : 1,200
Outline method 591 <0.5
survival of woodpeckers that operate
in such a way that the rotational
component of their pounding activity
is minimized.
Observations unrelated to head in¬
jury1014 suggest that the woodpecker
pecks in a straight line. If so, it would
be implicit that straight-pecking is
healthier—since the rotational compor
nents generated by successive flexion
and extension of the neck and by
rocking the body are being counter¬
acted by progressive extension of the
head during "strike" and by progres¬
sive flexion during "return," the bird
is presumably protected against rota¬
tional shearing injury. It also seems a
reasonable additional hypothesis that,
like certain caprini,läle the human
boxer,1 the gannet,1718 and the chim¬
panzee,8 the woodpecker might pro¬ 60 55 50 45 40 35 30 25 20 15 10 0 mm
tect itself further against rotational
and vibrational injury by holding its
Fig 2.—Woodpecker drilling trajectory. Hand plots of trajectory of center of eye during
neck rigid at the moment of impact.
(trajectory points numbered in 1 72,000 second).
four sequences
METHODS
To investigate these matters further, we rattle of typewriter keys would usually center of the eye could be defined most
examined color film of pileated woodpeck¬ suffice to elicit the pounding. clearly, and the outline was used to place
ers (Dryocopus pileatus) taken in nature For the color film, frame-by-frame the center of the eye. This involved error
(24 frames per second step-printed to slow display on a flat bed editing system was due to the grainy quality of the film and to
up the action). This suggested that our used to plot the trajectory of three points: alterations in neck outline at different
hypotheses might be correct, but the the tip of the beak, the cranial vertex, and points of the trajectory. (3) In a less subjec¬
camera speed was too slow to follow the the center of the head. tive computer analysis, the frames were
movements with certainty. Accordingly, For the black-and-white film, selected digitized using a microdensitometer set to
we made high-speed films (400 frames per interval frames were printed and the a 35-u scanning aperture, converting each
second in color and 2,000 frames per second trajectory of the center of the eye was into 152 x 230-element matrix of film
a
in black and white) featuring a mature plotted as follows: (1) Two judgments by density. A point corresponding roughly to
female acorn woodpecker (Melanerpes for- the same person, based on direct visual the center of the bird's head was located on
micivorus) that, having residual difficulty inspection, as to the location of the center the first frame of each sequence and
from a broken wing, was unable to fly and of the eye were averaged. This involved tracked from frame to frame. The location
lived free, uncaged, and still pounding error due to the grainy quality of the film of the point in a frame was predicted from
vigorously away in a tree trunk in the and to closure of the eyelid immediately its position in the immediately preceding
office of the park ranger at the Placerita before impact. (2) An outline of the head frame and its velocity through the two
Canyon State Park, Newhall, Calif. A was traced from frames in which the preceding frames. The error of prediction

Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 05/31/2015


- .'
30 35 40 45 50
Fig 3.—Woodpecker drilling trajectory. Computer plots of criterion point during
sequences(trajectory points numbered in 1 /2.000 second).
Table 2.—Drilling Impact: Computer
Maximum Velocity,
cm/s
Sequence 1 749
Sequence 2 623
due to acceleration was then corrected by
searching the vicinity of the predicted loca¬
tion with a template taken about the point
from the preceding frame. The goodness of
fit of the template was then calculated for
points in the vicinity, and the best-fit
location of the template was taken as the
true position. Error due to digitization was
reduced by interpolating the exact location
of the best fit; the interpolated value was
then used to calculate the exact velocity of
the criterion point.
RESULTS
Color Films
In all the sequences filmed, the
strike trajectory of the tip of the beak
was straight. The trajectories of the
vertex and the center of the head were
also essentially linear (Fig 1). Deceler¬
ation occurred in one frame (1/400
second) or less. A plot from a sample
strike sequence indicated a terminal
velocity of at least 382 cm/s and an
impact deceleration of at least 155 g.
As an interesting aside, we noted
that the woodpecker set about its
work in much the same manner as the
human carpenter. Before winding
back to strike in earnest, it often did
two preliminary light taps on the
target, presumably "to line it up."
Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 05/31/2015
I I I I I I I I I I i and 1,260 to 2,170 g (mean, 1,715 g) by
55 60 65
Y Computer Analysis.—Figure
Plots
Stopping Time,
Deceleration,
ms
£0.5_>1,525
. >
£0.5_>1,267
£1.0
This particular aspect of the wood¬
pecker's behavior has not, to our
knowledge, been described before.
We also noted that the eyelid shut
immediately before impact, then
opened immediately, and remained
open all the way back; it closed each
time at exactly the same point in the
strike phase-approximately 0.0025 s
(one frame) before impact. This clos¬
ing has been described previously, but
not its timing sequence. The mecha¬
nism is obscure. We speculate that, in
addition to keeping chips out of the
eye, as suggested by Spring,10 eyelid
closure could serve another important
function—a "seat belt" to keep the
eyeball from popping out on impact.
Black-and-White Film
Figure 2 shows hand plots of the
trajectory of the center of the eye
during four drilling sequences. The
outline method gives more plausible
plots than simple inspection, as well as
more conservative estimates of veloci¬
ty and deceleration (Table 1).
As plotted by the outline method,
the trajectory in sequences 1 and 2
was essentially linear with a very
slight, superimposed downward cur¬
vature. In sequences 3 and 4, the head
entered the hole, making it impossible
to plot the final frames; however, they
are a useful check on sequences 1 and
2, confirming the linear trajectory.
The maximum velocity in sequence
1 748 cm/s; and in sequence 2, 591
was
cm/s (mean, 679 cm/s). The inspection
method gives higher velocities, 1,060
and 618 cm/s, respectively. In the
incomplete sequences 3 and 4, it was
531 and 728 cm/s, respectively, clearly
of the same order.
The head stopped abruptly in one or
two frames, indicating a deceleration
time of 0.5 to 1.0 ms and an estimated
impact deceleration of 763 to 1,200 g
(mean, 982 g) by the outline method
70 the inspection method.
3 plots
the terminal trajectory of the criteri¬
on point in two sequences, and Table 2
two gives the relevant velocity and impact
data, all very much in line with the
hand plots.
The trajectories are basically
straight with maximum preimpact
velocities of 749 and 623 cm/s, respec¬
g_ tively (mean, 686 cm/s). In both
sequences, the head stopped in one
762 frame on the X axis and in two frames
on the Y axis. The Y movement in the
second frame is small and might be
> 634
attributed to residual error; it may,
however, indicate sliding movement
of the beak as it gouges out a wood
chip. If the deceleration time is taken
as 0.5 m/s, the mean impact decelera¬
tion was 1,396 g (range, 1,267 to 1,525
g). If the deceleration time is taken as
1.0 m/s, the mean impact deceleration
was 698 g (range, 634 to 762 g). There
was amazingly little movement after
the impact, and it is clear that the
head must have been held rigidly in
position at the time, presumably by
muscular contraction. (Finer vibration
within the millisecond time frame
may, of course, have occurred.)
COMMENT
Obviously, findings should not
our
be used as a basis for conclusions
about all woodpeckers, let alone all
head injury. Nevertheless, our wood¬
pecker withstood repeated impact
forces of the order of 1,000 g—a
formidable achievement, even if the
two-thirds power hypothesis of scale
and its attached package of assump¬
tions are accepted in toto. The lineari¬
ty of the head trajectory suggests that
during pecking the brain is virtually
free of rotational forces that might
impinge injuriously on it. That this
linearity of trajectory is indeed adap-
 tive and protective is implicit in the
fact that, to achieve it, a much more
complex neuromuscular mechanism
must undoubtedly intervene than if
pecking were performed with arcuate
head movements. In arcuate move¬
ments, whether the pivot is in the
neck or elsewhere, the alternating
contraction (and reciprocal relaxation)
of certain flexor and extensor mus¬
cles, at a frequency no greater than
the pecking frequency, would be suffi¬
cient. A linear trajectory, however,
necessitates a much finer and graded
adjustment of the whole postural
apparatus, carried out continuously at
high speed and involving a much more
elaborate integration of both extero-
ceptive and interoceptive informa¬
tion.
There is no "prima facie" evidence
that a linear trajectory is mechanical¬
ly more efficient for a muscle-
powered drill than an arcuate mo¬
tion—in fact, the evidence is to the
contrary, as an attempt to strike a
straight-line blow with a hammer will
demonstrate. It would seem, there¬
fore, that the justification for a more
complex, less economical mechanism,
requiring such rapid muscular adjust¬
ments and feedback data processing,
must lie in its protective value. The
evidence already cited for the deleter-
1. May PRA, Fuster JM, Newman P, et al:
Woodpeckers and head injury. Lancet 1:454-455,
1976.
2. Becher F: Untersuchungen an spechten zur
frage der funktionellen anpassung an die
mechanische belastrung. Z Naturforsch 8:192\x=req-\
203, 1953.
3. Ommaya AK, Hirsch AE, Martinez J: The
role of whiplash in cerebral concussion, in:
Proceedings of the Tenth Stapp Car Crash
Conference. Warrandale, Penn, Society of Auto-
motive Engineers, pp 197-203, 1966.
4. Unterharnscheidt F, Higgins LS: Pathomor-
phology of experimental head injury due to rota-
tional acceleration. Acta Neuropathol 12:200-204,
1961.
5. Goldsmith W: The physical processes pro-
ducing head injuries, in Caveness WF, Walker
AE (eds): Head Injury Conference Proceedings.
Philadelphia, JB Lippincott Co, 1966, pp 350\x=req-\
382.
6. Holbourn AHS: Mechanics of head injuries.
Lancet 2:438-441, 1943.
Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 05/31/2015
ious effects of rotational forces on
nervous tissue indicates how decisive
the protective value of linear pecking
must be for our bird.
It would seem prudent, therefore, to
pay attention to findings from study
of the morphology and dynamics of
the woodpecker, particularly since the
implications for head injury preven¬
tion are thoroughly in line with
modern packaging technology and the
results of animal experimentation and
mathematical modeling. The conver¬
gence of evidence from various
sources suggests that helmets that
claim to protect against anything
except a simple penetrating injury
should be thicker and, if possible,
lighter, and that they should be form-
fitted with a thick, firm, inner spongy
layer to absorb the energy of the
impact and a thin, harder, outer layer
to distribute the impact and resist
penetration and abrasion.
It is also necessary to discard the
magical notion that wearing a helmet
on the head is sufficient to protect
against impact brain damage. Atten¬
tion must be given to developing
systems that will dampen sudden
rotatory movements that could create
shearing strains in brain tissue; the
high collars of old fashioned military
uniforms and the vertex to chest- and
References
7. Liu YK, Chandrun KB, Von Rosenburg DU:
Angular acceleration of a viscoelastic (Kelvin)
material in a rigid spherical shell\p=m-\arotational
head injury model. J Biomech 9:285-292, 1975.
8. Ommaya AK, Hirsch AE: Tolerances for
cerebral concussion from head impact and
whiplash in primates. J Biomech 4:13-21, 1971.
9. Ommaya AK, Hirsch AE, Harris E, et al:
Scaling of experimental data in cerebral concus-
sion in sub-human primates to concussion thresh-
olds, in Proceedings of the 11th Stapp Car Crash
Conference. Warrandale, Penn, Society of Auto-
motive Engineers, 1967, pp 47-52.
10. Spring LW: Climbing and pecking adap-
tions in some North American woodpeckers.
Condor 67:457-488, 1965.
11. Caffey J: On the theory and practice of
shaking infants. Am J Dis Child 124:161-169,
1972.
12. Caffey J: Significance of the history in the
diagnosis of traumatic injury to children. J Pedi-
atr 67:1000-1014, 1965.
13. Caffey J: Multiple fractures in the long
shoulder-casing of the medieval
knight may have had some value in
this respect.1 Whenever feasible and
particularly in occupations or emer¬
gency situations that place subjects at
special risk (eg, car racing, stunts,
aircraft crashes, and pilot ejections),
the head injury system should restrain
or restrict whiplash rotation of the
head on the neck and of the neck on
the trunk. When the possibility of
imminent injury can be anticipated,
the persons at risk should be taught to
tighten the neck muscles and to hold
the neck flexed in chin-down position.
This has been shown to minimize
cranial rotation and to decrease peak
angular velocity and angular accelera¬
tion.19
This investigation was supported in part by the
Research Service of the Veterans Administra¬
tion.
Thomas R. Howell, Department of Biology,
University of California at Los Angeles, encour¬
aged our research. Use Lowenstam translated
Becher's article. Franz Steid and Paul Newman
did the illustrations. The color film of woodpeck¬
ers in nature was provided by Walt Disney
Productions, Burbank, Calif. The high-speed film
of the acorn woodpecker was taken by Steven
MacLain. V. D. Dhir, Department of Engineer¬
ing, University of California at Los Angeles,
provided the high-speed camera. Milton Plesset,
California Institute of Technology, assisted us
through an unfamiliar methodology.
bones of infants suffering from subdural hema-
toma. Am J Roentgenol Radium Ther Nucl Med
54:163-173, 1946.
14. Bock WJ: Kinetics of the avian skull.
J Morphol 114:1-41, 1964.
15. Reed CA, Schaffer WM: How to tell the
sheep from the goats. Field Museum Natural
History Bull 43:2-7, 1972.
16. Schaffer WM, Reed CA: The co-evolution
of social behavior and cranial morphology in
sheep and goats. Fieldiana 61:187, 1972.
17. Gordon D: Woodpeckers, gannets, and head
injury. Lancet 1:801-802, 1972.
18. May PRA, Fuster JM, Newman P, et al:
Woodpeckers and head injury. Lancet 1:1347\x=req-\
1348, 1976.
19. Ewing CL, Thomas DJ, Lustick L, et al:
The effect of the initial position of the head and
neck on the dynamic response of the human head
and neck to -Gx impact acceleration, in Proceed-
ings of the 19th Stapp Car Crash Conference.
Warrandale, Penn, Society of Automotive Engi-
neers, 1976, pp 487-512.