Anatomical Predictions of Hearing in
the North Atlantic Right Whale
SUSAN E. PARKS,* DARLENE R. KETTEN, JENNIFER T. O’MALLEY,
AND JULIE ARRUDA
Woods Hole Oceanographic Institution, Woods Hole, Massachusetts
Some knowledge of the hearing abilities of right whales is important
for understanding their acoustic communication system and possible
impacts of anthropogenic noise. Traditional behavioral or physiological
techniques to test hearing are not feasible with right whales. Previous
research on the hearing of marine mammals has shown that functional
models are reliable estimators of hearing sensitivity in marine species.
Fundamental to these models is a comprehensive analysis of inner ear
anatomy. Morphometric analyses of 18 inner ears from 13 stranded North
Atlantic right whales (Eubalaena glacialis) were used for development of
a preliminary model of the frequency range of hearing. Computerized to-
mography was used to create two-dimensional (2D) and 3D images of the
cochlea. Four ears were decalcified and sectioned for histologic measure-
ments of the basilar membrane. Basilar membrane length averaged
55.7 mm (range, 50.5 mm–61.7 mm). The ganglion cell density/mm aver-
aged 1,842 ganglion cells/mm. The thickness/width measurements of the
basilar membrane from slides resulted in an estimated hearing range of
10 Hz–22 kHz based on established marine mammal models. Additional
measurements from more specimens will be necessary to develop a more
robust model of the right whale hearing range.
? 2007 Wiley-Liss, Inc.
Anat Rec, 290:734–744,
Key words: cetacean; hearing;cochlea; basilarmembrane;
North Atlantic right whales (Eubalaena glacialis) are
among the most endangered mysticetes (baleen whales)
in the world. Although these whales have been protected
from whaling for much of this century, today fewer than
350 remain (Hamilton and Martin, 1999; Knowlton and
Kraus, 2001). Over 30% of known right whale mor-
talities in the past 30 years have been attributed to
collisions with vessels or entanglement in fishing gear
(Kraus, 1990; Kenney and Kraus, 1993; Laist et al.,
2001). Reducing human contributions to mortality is im-
perative for survival of this species. At present, we do
not know whether the whales are able to hear or localize
an approaching vessel. Failure to hear vessels could be
due to limitations of hearing in right whales or acoustic
propagation anomalies (e.g., the Llyod mirror effect;
Richardson et al., 1995) near the surface. One method
proposed to reduce vessel strikes is to equip ships with
alarm devices that alert whales to an approaching vessel
(Nowacek et al., 2003). Similarly, development of right
whale targeted acoustic pinger systems for fishing gear
could reduce mortality from entanglement. Thus, infor-
mation about the frequency range and sensitivity of
hearing in right whales is necessary to determine
whether hearing limitations play a role in vessel colli-
sions, to identify appropriate frequency bands for acous-
tic pinger stimulus development, and to determine what
anthropogenic noise sources may be affecting hearing
and acoustic communication in this species. This study
*Correspondence to: Susan E. Parks, Pennsylvania State Uni-
versity, Applied Research Laboratory, P.O. Box 30 State College,
PA 16804-0030. E-mail: email@example.com
Received 2 March 2007; Accepted 6 March 2007
Published online in Wiley InterScience (www.interscience.wiley.
Grant sponsor: The Northeast Consortium; Grant numbers:
02-560and 03-675;Grant sponsor:
through the Right Whale Initiative of the WHOI Ocean Life
? 2007 WILEY-LISS, INC.
THE ANATOMICAL RECORD 290:734–744 (2007)
uses anatomical measurements to predict the hearing
abilities of right whales.
HEARING STUDIES IN CETACEANS
Most hearing data from cetaceans come from studies
with small captive toothed-whales. Audiograms have
been made from 10 of these species (Au, 2000). Right
whales are not amenable to conventional behavioral or
electrophysical methods for measuring hearing because
of their large size and endangered status. The frequency
range of best hearing is commonly thought to overlap
with that of stereotypical species calls (Sales and Pye,
1974). Most baleen whale vocalizations have the major-
ity of their energy below 1 kHz. The average ambient
noise levels in all 1/3-octave bands below 1 kHz are
higher than 75 dB re 1 mPa (Urick, 1983). Ambient noise
levels may limit the real world detection threshold for
baleen whales. In the field, information on baleen hear-
ing abilities in the context of ambient noise can be
obtained from behavioral responses of whales to play-
back stimuli. Playback experiments with several species
have indicated good directional hearing capabilities in
baleen whales based on orientation toward and localiza-
tion of conspecific calls (Clark and Clark, 1980; Watkins,
1981; Tyack, 1983; Parks, 2003) and clear responses
of gray whales (Eschrichtius robustus) to the calls of
killer whale predators (Orcinus orca) (Cummings and
Thompson, 1971). Studies of baleen whale response to
anthropogenic noise sources documented response at fre-
quencies up to at least 15 kHz (Watkins, 1986). One play-
back experiment estimated the broadband received level
of sound necessary to elicit an approach response from
humpback whales (Megaptera novaeangliae) to be 102 dB
re 1 mPa for feeding sounds (Frankel et al., 1995).
ANATOMICAL MODELING FOR MARINE
Comparative anatomical studies have identified struc-
tural correlates to frequency range and hearing sensitiv-
ity in multiple mammalian species (Echteler et al.,
1994). Functional studies of the inner ear focus on reso-
nance characteristics of the basilar membrane. Models
have been developed to predict the frequency range of
hearing for cetaceans in particular based on basilar
membrane measurements (Ketten and Wartzok, 1990).
Position–frequency maps and basilar membrane elastic-
ity measurements from humans as well as other mam-
mal and bird species were used by Greenwood to derive
formulae for predicting frequency maxima, minima, and
distribution along the length of the basilar membrane
for land mammals (Greenwood, 1961, 1962, 1990). Other
estimates of overall hearing ranges have been based on
either cochlear length or length and width (Manley,
1971; West, 1985). Fay’s (1992) extrapolation of Green-
wood’s work shows that estimators derived from even a
single basilar membrane dimension provide very close
approximations of psychophysical measures of hearing
for most land mammals, referred to as generalists. How-
ever, these formulae provide a poor fit to hearing curves
of species with specialized hearing, e.g., the horseshoe
bat (Rhinolophus ferrumequinum) and the mole rat
(Spalax ehrenbergi), largely because these species have
a basilar membrane stiffness gradient that differs from
the generalist basilar membrane (Echteler et al., 1994).
Species with high frequency hearing tend to have com-
paratively narrow and thick basilar membranes with
better developed outer spiral laminae that continue
through most of the cochlear duct (Ketten, 1984). In spe-
cies with ears specialized for low frequency hearing, the
basilar membrane is generally wider and thinner, with
the outer spiral laminae thinner and if present, located
only in the basal region of the cochlea (Echteler et al.,
1994). Measurements of basal and apical stiffness of the
basilar membrane, which can be approximated by the
ratio of thickness to width, appear to predict the upper
and lower frequency limits equally well for both general-
ists and specialist ears (Ketten and Wartzok, 1990), and
comparative anatomical studies of cetacean and terres-
trial mammalian ears demonstrate significant structural
variants unique to cetacean ears (Ketten, 1992).
Whale ears have the same basic components as land
mammal ears but they also have adaptations to the
aquatic environment that require more comprehensive
modeling. Generalized morphometric models for land
mammals provide a procedural or mechanistic basis for
marine mammal analyses, but these must be modified to
accommodate structural differences in whale ears com-
pared with those from typical land mammals. Whale
basilar membrane thickness and width differ from that
typical of land mammal ears, and consequently, their
basilar membranes do not follow the generalist stiffness
gradients for their membrane lengths. Therefore, esti-
mates using generic land mammal formulae based on
length alone are incorrect for cetaceans (Ketten, 1984).
The appropriateness of a more comprehensive model
for whales was first demonstrated in a structural analy-
sis of the cochlea of 12 odontocete species (Ketten, 1984;
Ketten and Wartzok, 1990). The results of these studies
showed that a combination of four measurements of
cochlear structure (basilar membrane dimensions, lami-
nar extent, membrane pitch, and basal turn ratio)
allowed for excellent prediction of the primary bands of
ultrasonic hearing in odontocetes (Ketten, 1984; Ketten
and Wartzok, 1990).
This study describes the anatomy of right whale ears
and uses the model (Ketten, 1994) based on the mea-
surements of the basilar membrane pitch, basal turn
ratio, and basilar membrane dimensions to estimate
their frequency range of hearing. The anatomy of the
right whale ear indicates that they are specialized for
low frequency hearing.
MATERIALS AND METHODS
Specimen Collection and Preservation
The endangered status of the North Atlantic right
whale has led to systematic necropsies on all recovered
dead right whales (Moore et al., 2004). Right whale
temporal bones have been routinely collected from
necropsies since 1989. Ears were either frozen shortly
after collection or placed in a buffered 10% formalin
solution. Preservation condition ranged from code 2 to
code 4 (2, fresh; 3, decomposed; 4, severely decomposed)
(Moore et al., 2004). A total of 18 ears were analyzed
from 13 different individuals (Table 1). All ears were
computer tomography (CT) scanned, and four ears were
further processed into slides for histology.
NORTH ATLANTIC RIGHT WHALE HEARING
Each specimen was imaged using CT scanning. The
specimens were scanned with a Siemens Spiral Plus 4
CT (Massachusetts Eye and Ear Infirmary) or a Siemens
Emotion Spiral CT and Volume Zoom scanner (Woods
Hole Oceanographic Institution). Scans were obtained
with a 1-mm spiral ultra–high-resolution protocol and
reconstructed at 0.5-mm slice thickness. Three speci-
mens were additionally scanned using a 0.5-mm spiral
acquisition. Both 2D and 3D representations of the scan
data were used for measuring the cochlear anatomy of
the right whale ears (Figs. 1–4).
The number of turns in each cochlea was determined
from 3D reconstructions of the cochlear duct oriented for
a top-down, apex to base, view of the cochlea (Fig. 3B).
Cochlear lengths were determined by measuring the
length of cochlear turn radii for multiple positions in the
cochlea from 2D paramodiolar (Fig. 4) cross-sections.
The paramodiolar slices were formatted to be perpendic-
ular to the longer axis of the basal turn. These values
were then used to calculate the axial pitch, basal ratio,
and the length of the cochleae. The length calculations
were made using the following formulae (Ketten et al.,
(1) For r ¼ au
where z ¼ cochlear length, r ¼ radius at angular dis-
placement y in radians, a ¼ constant that determines
the size of the spiral, and h ¼ axial height of the spiral
The 3D reconstructions of each cochlea were measured
directly to compare the observed cochlear length from
the CT scan with calculated cochlear length based on 2D
radii measurements. The basal diameter of the cochlea
was measured from the 3D reconstructions in the same
orientation as the radial cross-sections. Small differences
among these measures are to be expected because the
TABLE 1. Specimens measured for this studya
Specimen StrandingAgeSex Length (cm)Code PreservationCT Histology
aData given indicate the date of the stranding, the age (calf < 6 months in age, adult > 8 years
of age) of the whale at time of death, sex, total body length measured from snout to fluke notch,
state of preservation of the specimen at time of necropsy (Code: 1 ¼ best, 4 ¼ worst), and
whether the specimen was analyzed using CT and histology. Asterisks indicate that both ears
were analyzed from these specimens.
men EG 5. The reconstruction is from 0.5-mm sections of computed tomography scans. The tympanic
and periotic are labeled in each image. A: Medial view. B: Anterior view. C: Lateral view.
A three-dimensional (3D) reconstruction of the entire left temporal bone complex from speci-
PARKS ET AL.
3D images are oriented as flat projections and include
the basal hook of the cochlear spiral, while the calcu-
lated lengths include the cochlear rise but do not include
the basal hook length.
Gross Dissection and Histology Measurements
Specimens with evidence of preservation of the vesti-
bulocochlear (VIIIth cranial) nerve and of the basilar
membrane based on the CT scan images were selected
for further processing. These specimens were dissected,
which involved defrosting frozen specimens in 10% buf-
fered formalin solution and removal of all remaining
external soft tissue (Fig. 6). The remaining tissues were
measured and weighed. The periotic and tympanic bones
were then separated and the bony flanges were removed
from the periotic by use of a handsaw to reduce the
volume of bone surrounding the cochlea before deca-
lcification. When present, ossicles were removed and
preserved in a 1% formalin solution for use in density
measurements at a later date.
The periotic bones from the gross dissection were
placed into solution to decalcify the hard tissue sur-
rounding the cochlea to allow sectioning for microscopy.
The ears were decalcified in 5% trichloroacetic acid,
ethylenediaminetetraacetic acid (EDTA), or in acid with
later transfer to EDTA. Bone wax was placed in the oval
and round windows of specimens decalcified in acid to
reduce the impact of the acid on the soft tissues of the
cochlea. Acid decalcification of various durations were
attempted to determine whether this method is accepta-
ble to decrease time for decalcification and to determine
the effects of acid techniques on inner ear structures,
especially the basilar membrane.
After decalcification was complete, the ears were
embedded in celloidin solution to harden and sectioned
into 20-mm sections. All sections were retained and every
10th section was stained with hematoxylin and eosin
and mounted as cover-slipped slides (e.g., Fig. 7).
Basilar membranes that were present in the slide sec-
tions were measured for width and thickness (Fig. 8).
The width was measured at a 403 objective magnifica-
tion on a light microscope (Olympus Model BX40) with a
graticule and ocular (103) calibrated scale for width
measurements. Oil immersion microscopy using a 1003
oil immersion objective was used to measure basilar
membrane thickness. Reconstructions of the basilar
membranes were made using measurements of the basi-
lar membranes from the slides and fitting them to an
equiangular spiral with the same (a) value calculated for
chain from a right whale. The figure shows the in situ orientation, the
angle between ossicles, as well as the position of the ossicles relative
to the cochlea. A cross-section through the cochlea can be seen on
the right side of the figure. This cross-section is approximately the
same orientation as the cross-section shown in Figure 4.
A three-dimensional (3D) reconstruction of the ossicular
showing the height of the right whale cochlea spiral and the VIIIth cranial nerve. B: View of the cochlear
spiral illustrating the number of turns in the right whale cochlea, with 2.3 turns present in this specimen.
A three-dimensional (3D) reconstruction of the cochlea from specimen EG 6. A: Lateral view
NORTH ATLANTIC RIGHT WHALE HEARING
the right whale ears. Figure 9 shows the measurements
of intact basilar membranes fit onto a curve.
Using the measured thickness and width of the basilar
membrane at multiple positions along its length, the
absolute and functional hearing range of the right whale
can be estimated using the model described originally
for odontocetes (Ketten and Wartzok, 1990) that is now
conventionally used for human CT data (Ketten et al.,
1998) and has been applied to mysticetes (Ketten,
1994). The functional hearing range is generally some-
what narrower than the total possible inner ear re-
Ganglion cells were counted from each slide from all
sectioned specimens, and total number of ganglion cells
and total number of ganglion cell nuclei were counted at
403 resolution using a grid. The number of ganglion
cells counted from the mounted slide sections were mul-
tiplied by 10 (to account for the unmounted sections)
(Schuknecht, 1993). The Konigsmark correction was
used to avoid double counts from cells split between sec-
tions (Nadol, 1988: Ncorr¼ Ncoun[t /t þ d] where Ncorr¼
corrected cell count, Ncoun¼ actual cell count, t ¼ thick-
ness of the section and d¼ diameter of the cell counted).
In this case, the thickness of the section ¼ 20 mm and
the diameter of the cells is approximately 20mm (range,
15 mm to 25 mm), leading to a correction factor of 0.5.
All specimens show signs of decomposition and loss of
ganglion cells. Therefore, combined counts pooling the
highest ganglion cell counts for a given basilar mem-
brane position from all specimens were used to estimate
the total number of ganglion cells present in a right
whale ear. The ganglion cell density/mm was calculated
as a percentage of the basilar membrane length for the
two specimens with the best preservation. The density
counts from these two specimens were also pooled to get
a better estimate of the density of ganglion cells per unit
length of the basilar membrane.
Initial surveys of cochlear dimensions from CT images
showed that precise orientation of the cross-sections
taken by the CT scanner is important for consistent
measurement of all cochlear features of right whale
cochlea (Table 2). The calculated cochlear length and the
cochlear length measured directly from 3D CT recon-
structions are shown in Table 3. There was good agree-
ment between the predicted and observed cochlear
length for most specimens. However, the values meas-
ured from the CT images are slightly longer. The added
length is attributable to both minor shrinkage in
sectioned material and to inclusion of the terminal hook
in the measurements from the CT images.
There were several aspects of the temporal bone anat-
omy seen in all dissected specimens that are noteworthy.
First, the juncture between the periotic and tympanic
bones is extremely stable, with 2- to 5-mm thickness of
bone providing two points of connection. The largest con-
tact point, directly lateral of the cochlea itself, was often
1–2 cm in length. The second connection, at the anterior
end of the periotic, was generally smaller and thinner.
These two connections formed an arch between the
periotic and tympanic through which the ‘‘glove finger’’
A: Schematic showing the orientation of the cross-section. B: An example of the resulting two-dimen-
sional CT image from the cross-section.
Cross-section through a right whale cochlea for computerized tomography (CT) measurements.
sional computerized tomography cross-sections from right whale
specimens with 2.5 turns.
Illustration of the measurements made from two-dimen-
PARKS ET AL.
(a derivative of the tympanic membrane) projected later-
ally (Fig. 6).
The entire ossicular chain was present in most speci-
mens (Fig. 2). The malleus was supported by a bone
strut that connected to the outer wall of the tympanic
bone, close to the insertion of the glove finger. The
corpus cavernosum covered the interior of the tympanic
bone in three of four specimens. A spongy layer of sharp
spicules of cancellous bone and fatty tissue covered the
dorsal–lateral side of the periotic bone. This spongy
layer was very difficult to remove from the specimens,
and it covered a dome of very dense bone in the periotic
that projected laterally from the side of the cochlear
duct itself. This very dense layer of bone was one of the
major roadblocks to decalcification in these specimens.
This dense bone and spongy pad correspond to a position
immediately between the bony flanges that wedge the
ears against the skull.
Acid decalcification produced multiple artifacts. Mid-
modiolar cross-sections show the impact of the acid decal-
cification (Fig. 10). Comparing the preparations of EG 1
medial view of the right ear. The tympanic and periotic bones are
labeled. The VIIIth cranial nerve canal, round window, and bony flange
are also labeled. B: The lateral view of the same ear. The tympanic,
Images from dissection of EG 11 temporal bones. A: The
periotic, and flange are labeled to aid in orientation. The glove finger,
which is the common term for the baleen whale tympanic membrane,
is in its normal position between the two temporal bone elements.
9. A: A mid-modilar cross-section showing the layout of a right whale
cochlea. The basal and apical turns are labeled. The VIIIth cranial
nerve is also labeled. The basal turn is marked by a white square. B:
The basal turn from under higher (153) magnification. The basilar
Images from histology slide preparations from specimen EG
membrane, spiral ligament, and outer osseous spiral lamina are la-
beled. Many inner ear structures are absent (e.g., Reissner’s mem-
brane separating the scala vestibuli and scala media) as a result of
postmortem changes and histology preservation and decalcification.
NORTH ATLANTIC RIGHT WHALE HEARING
(no acid) (Fig. 10A) and EG 18 (only acid) (Fig. 10D)
illustrates the effect of acid decalcification. There is clear
decalcification of the bone surrounding the cochlea from
EG 18, and all of the inner osseous spiral laminae
have been dissolved, resulting in total disruption of any
basilar membrane that may have remained when the
specimen was collected.
Ganglion cells. The preservation of ganglion cells
varied in the specimens sectioned for slides. The total
corrected ganglion cell count for the two specimens with
the best preservation are 37,930 for EG 4 and 31,390 for
EG 9. This finding represents a minimum estimate of
cells as there is clear evidence of neuronal loss from dis-
ease and/or decomposition in these specimens. A count
of 45,250 is obtained by combining the highest ganglion
cell count for a particular position on the basilar mem-
brane from EG 4 and EG 9. Calculation of the ganglion
cell density/mm results in an average value of 1,842
ganglion cells/mm (Table 4), which is likely to be a more
measurement of the width of the membrane. The thickness of the
membrane was measured in the center of the membrane.
Basilar membrane from EG 9 at 203 marking points for
ments. The gray bars represent the length of the basilar membrane at particular points on the cochlea.
The black numbers represent the slide number. The light gray numbers are the count of ganglion cells at
each point on the corresponding slide.
Image of a reconstructed basilar membrane with ganglion cell counts from the slide measure-
PARKS ET AL.
TABLE 2. Radii measurements taken from the computed tomography scans for all specimensa
1/2 p 3/2 p 5/2 p7/2 p 9/2 p
aThe radii (1/2 p, 3/2 p, 5/2 p, 7/2 p, and 9/2p for 2.4–2.5 turns, 0, p, 2p, 3p, and 4 p for 2.25 turns), axial height and basal
diameter are reported in millimeters. The number of turns is derived from counts made on three-dimensional reconstruc-
tions of the cochlea for each specimen. The axial pitch ¼ axial height/number of turns, and the basal ratio ¼ axial height/
basal diameter defined by Ketten (1984).
TABLE 3. Measurements used for the calculation of length of the cochlear canala
Theta Axial height
a ¼ Spiral
z ¼ Calculated
cochlear length (mm)
aTheta is the number of degrees in the spiral reported in radians. Axial height is the height of the
spiral in millimeters. The spiral constant (a) is calculated to give the relative size of the approxi-
mated spiral from formula (1). Calculated cochlear length (z) is calculated from formula (2).
A: Ear from EG 1 (calf) decalcified in ethylenediaminetetraacetic acid
(EDTA) only. B: Ear from EG 4 (calf), 2 months in trichloroacetic acid,
14 months in EDTA. C: Ear from EG 9 (calf), 1.5 months in trichloro-
acetic acid, 5 months EDTA. D: Ear from EG 18 (adult), 6 months in
Impact of acid decalcification on baleen whale ear bones.
acid decalcification with bone wax in the oval and round windows to
reduce the time the tissues of the inner cochlea were exposed to
acid. The start of the second turn is circled in each slide for compari-
son of the acid effects on the inner osseous spiral lamina. The scale
bar in each image ¼ 1 mm.
NORTH ATLANTIC RIGHT WHALE HEARING
representative estimator of total population, when multi-
plied by length, than are the partial totals obtained in
these tissues. Combined with the average cochlea length
(55.7 mm, Table 3), the probable total ganglion cell num-
ber is approximately 102,500 ganglion cells, assuming
equal neuronal distribution throughout the length of the
basilar membrane. The lower actual counts in this study
are clearly the result of cellular loss from specimen
decomposition as well as possible in vivo pathologies.
Basilar membrane measurements and pre-
dicted frequency range of hearing. All specimens
had measurable intact membranes in some region of the
cochlea. The wider, thinner membranes near the apical
turn of the cochlea were often poorly preserved, while
the shorter, thicker portion of the membrane supported
by the outer osseous spiral laminae was intact for all
specimens. Table 5 shows the thickness/width ratios at
different sections of the membrane length and the esti-
mated frequency range of hearing for the right whale
calculated from the data using the model described in
Ketten (1994). These numbers represent the functional
range of the average of the ears measured in this study.
These values also indicate a total possible hearing range
of the right whale of approximately 10 Hz–22 kHz.
The results of this study provide a description of
North Atlantic right whale ear morphology and mea-
surements of their cochlear and basilar membrane
dimensions. The highly endangered status of the North
Atlantic right whale has resulted in concerted efforts to
perform complete necropsies on all dead right whales
that can be recovered. This has resulted in the collection
of baleen whale ears that are in good condition when
from relatively fresh specimens. Better specimens could
be collected from fresh kills in populations of baleen
whales where whaling still occurs. This is not an option
for any right whale population anywhere in the world.
This study indicates that collection of ear bones from
right whales in any state of decomposition is worth-
while. This may have great application to collection of
approaches are similarly limited. Ears collected from
highly decomposed specimens often retained the ossicu-
lar chain position in situ. Specimens with moderate
decomposition retained the tough glove finger and lining
of the tympanic chamber. All specimens allowed for mea-
surement of size, length, and number of turns in the
cochlea of the right whale. Decalcification and histologi-
cal processing of two Code 3 (moderate decomposition)
specimens resulted in very useful basilar membrane
measurements and ganglion cell counts. Specimens in
any condition from other right whale populations would
be of use to determine whether there are significant dif-
ferences in ear anatomy among the three proposed spe-
cies of right whales.
fromother speciesfor whichresearch
The CT scanning of specimens proved effective for a
variety of in situ measurements of right whale ears.
Both 2D and 3D images were useful in describing the
right whale cochlea. The 2D reconstructions were used
to evaluate the condition of the middle ear and to detect
any remnants of the VIIIth cranial nerve and in some
cases the basilar membrane condition to assist with
selection for further dissection and histological process-
ing. Cochlear length could be determined both from
measurement of radii from 2D cross-sections and from
direct measurement of 3D reconstructions of the canal.
The total number of turns was best determined from the
The dimensions, position, and orientation of other ear
structures are simultaneously available from CT scans
of intact ears, including the vestibular system and
ossicles (Fig. 2). Measures of these structures will allow
for future estimates of middle ear transfer functions for
this species. Scans of entire temporal complexes could be
used for measurement of the size of the tympanic
and the periotic bones. All of these observations can be
made without disturbing the positions or destroying the
Gross Dissection and Histology
The gross dissections of ears provided data relating to
middle ear structures in the bony ear complex of right
TABLE 4. Calculated ganglion cell density
per millimeter of the basilar membrane
Percent membrane length
TABLE 5. Thickness/width ratio of the basilar
membrane measurements and predicted frequency
response at different percentages of membrane
length (apex ¼ 0) combined from measurements
from four individual specimensa
(apex to base)
aThe thickness/width ratio was only measured from nontan-
PARKS ET AL.
whales. The spongy cancellous bone covering the densest
dome of bone of the periotic may provide insight into the
mode of sound transduction in right whales (see also
Nummela et al., 2007, this issue). If bone conduction of
sound is important for hearing in right whales, then the
flanges may function to direct the sound to the dense
bone surrounding the cochlea, with the spongy pad
reducing incoming bone conduction of sound from other
directions. Alternatively, the spongy pad of bone against
the very dense bone surrounding the cochlea could func-
tion to isolate the cochlea from vibrations of the skull.
The bony strut found supporting the ossicles and the
presence of a well-developed stapedial muscle indicate
that the ossicular chain in right whales may be func-
tional. Even with the strut and the muscle present, the
ossicular chain can be moved. Further mechanical stud-
ies need to be conducted to determine the functional role
of the ossicles in baleen whale hearing.
Decalcification times for right whale ears was longer
than would be predicted simply from the mass of the
ears, primarily because of the exceptional density of the
periotic, the bone surrounding the cochlea. Decalcifica-
tion of an isolated, pared down, periotic bone took from
6–20 months, depending on the size of the specimen and
the time in acid. The 5% trichloroacetic acid was used in
an attempt to accelerate the rate of decalcification. The
acid did substantially increase the rate of decalcification.
For example, EG 18’s left ear decalcified in 6 months
solely in acid, but EG 18’s right ear took 20 months in
EDTA. However, acid alone as a decalcificant created
multiple and significant artifacts, including the loss of
soft tissue within the cochlea (Fig. 10). In baleen whales,
the width and thickness of the basilar membrane toward
the apical turn is the most important feature for accu-
rate measurements to estimate low frequency hearing
sensitivity. Unfortunately, membranes in the apex are
the most fragile and the first lost in acid decalcification.
None of the specimens decalcified in acid retained any
basilar membrane beyond the first turn of the cochlea. A
chelating agent such as EDTA is far superior. EDTA
minimizes artifacts from decalcification and better pre-
serves fragile membranes (Schuknecht, 1993). There is
evidence that even specimens decalcified in EDTA can
be overdecalcified. The extremely dense section of bone
that projected laterally from the cochlear canal was the
last area to decalcify, and it is likely that the cochlear
canal itself was sufficiently decalcified to allow cutting
significantly earlier. Future attempts at decalcification
should focus on removing as much of the high density
periotic as possible before and during decalcification as
more is exposed to accelerate the process. Specimens
from calves and adults made it possible to observe aging
effects, such as demineralization of the periotic bone in
older whales. Figure 10a–c shows ears from calves (<6
months in age) while Figure 10d is an ear from an adult,
at least 23 years of age. The latter has lower bone den-
sity surrounding the cochlea. This change is not merely
a result of differences in decalcification of the preserved
specimen, but rather a real density difference that was
evident from CT scans before the ear was dissected and
The basilar membrane dimensions of the right whale
are consistent with previously described measurements
of baleen whale basilar membranes (Wartzok and Ket-
ten, 1999). The base of the basilar membrane is thicker
and narrower than the apical turn, which is extremely
thin and wide. The apical turn of the right whale ear
has membranes that may be thinner than can be accu-
rately measured by traditional light microscopy and are
perhaps best examined by transmission electron micros-
The distribution of ganglion cells in the best-preserved
specimens indicates that there may be variable ganglion
distribution and possibly hair cells in different regions
in the cochlea of right whales. However, the postmortem
decomposition of the specimens makes interpretation of
the remaining ganglion cells difficult. Combining the
counts of the best preserved sections of the basilar mem-
branes from EG 4 and EG 9 still yielded a total ganglion
cell count that was significantly lower than any reported
ganglion cell count for any cetacean species due to large
segments with total loss of cells. The direct counts are
slightly greater than seen in human ears (30,000) but
much less than half of what has been reported for other
baleen whales (156,000) (Ketten, 2000). The density of
cells in relatively well preserved areas was consistent
with previous cetacean ganglion cell data (Ketten, 2000).
The estimated ganglion cell densities/mm coupled with
the average length of the right whale basilar membrane
yielded a count comparable to those of other cetaceans.
It is notable that the ganglion cell count of the right
whale, presented here, and other baleen whales rival
those of odontocetes and these counts are much higher
than the average for any terrestrial mammal (Ketten,
The total hearing range for the right whale predicted
from measurements presented here is 10 Hz–22 kHz
with functional ranges probably being 15 Hz–18 kHz.
These estimates were made using the model described
in Ketten (1994). The model has been shown to accu-
rately predict the frequency range of hearing in both
odontocetes and bat species. Currently, there are no
direct measures of baleen whale hearing; therefore, the
results from this model cannot be compared with behav-
ioral or physiological hearing curves for right whales.
The robustness of this model makes it likely that the
frequency range of hearing presented here is a close
approximation to the hearing abilities of this species.
The apical measurements of the basilar membrane indi-
cate better low frequency hearing than in humans, while
the capacity suggested by the basal end of the mem-
brane is slightly higher in frequency but similar to
human ears. As expected, this range corresponds well to
the sounds produced by right whales (Parks and Tyack,
2005). Both this frequency range and the frequency
range of right whale sounds overlap with the frequency
range of many anthropogenic noise sources, suggesting
that noise could potentially have a negative impact on
This study represents a rare look at multiple ear
specimens from a single baleen whale population. It
provides a comprehensive description of multiple ear
specimens collected from an endangered baleen whale
species. It is difficult to collect baleen whale ear speci-
mens as large whale strandings are relatively rare in
comparison to those of small odontocetes. As expected,
there was variation in the size, length, and number of
turns of cochlea from different individuals, but consistent
intraspecies spiral form and length. Further research is
NORTH ATLANTIC RIGHT WHALE HEARING
needed to ground-truth these model predictions with Download full-text
field tests of the functional upper frequency of hearing
of right whales and the relative sensitivity of right whales.
The specimens used in this study were collected with
the assistance of numerous research staff and volun-
teers. Sincere thanks to everyone involved in the speci-
men collection, including M. Moore, the Cape Cod
Stranding Network, the Mid-Atlantic Stranding Net-
work, and the New England Aquarium right whale
research group. Members of the Ketten laboratory, par-
ticularly S. Cramer and J. Fenwick, at the Woods Hole
Oceanographic Institution provided support and assis-
tance with the project. P. Tyack supported S.E.P. during
her graduate work and provided constructive sugges-
tions and comments during all stages of this work.
S.E.P. was supported in part by a NDSEG Fellowship
and the Woods Hole Oceanographic Institution Educa-
Au WWL. 2000. Hearing in whales and dolphins: an overview. In:
Au WWL, Popper AN, Fay RR, editors. Hearing by whales and
dolphins. New York: Springer-Verlag. p 1–42.
Clark CW, Clark JM. 1980. Sound playback experiments with south-
ern right whales (Eubalaena australis). Science 207:663–665.
Cummings WC, Thompson PO. 1971. Gray whales, Eschrichtius
robustus, avoid the underwater sounds of killer whales, Orcinus
orca. Fish Bull 69:525–530.
Echteler SW, Fay RR, Popper AN. 1994. Structure of the mamma-
lian cochlea. In: Popper AN, editor. Comparative hearing: mam-
mals. New York: Springer-Verlag. p 134–171.
Fay RR. 1992. Structure and function in sound discrimination
among vertebrates. In: Popper AN, editor. The evolutionary biol-
ogy of hearing. New York: Springer-Verlag. p 229–267.
Frankel AS, Joseph R. Mobley J, Herman LM. 1995. Estimation of
auditory response thresholds in humpback whales using biologi-
cally meaningful sounds. In: Kastelein RA, Thomas JA, Nachti-
gall PE, editors. Sensory systems of aquatic mammals. Woerden,
The Netherlands: De Spil Publishers. p 55–70.
Greenwood DG. 1961. Critical bandwidth and the frequency coordi-
nates of the basilar membrane. J Acoust Soc Am 33:1344–1356.
Greenwood DG. 1962. Approximate calculation of the dimensions of
traveling-wave envelopes in four species. J Acoust Soc Am
Greenwood DG. 1990. A cochlear frequency-position function for
several species-29 years later. J Acoust Soc Am 87:2592–2605.
Hamilton PK, Martin SM. 1999. A catalog of identified right whales
from the Western North Atlantic: 1935 to 1997. Boston: New Eng-
land Aquarium, Central Wharf. 27 p þ 382 plates.
Kenney RD, Kraus SD. 1993. Right whale mortality - a correction
and an update. Mar Mammal Sci 9:445–446.
Ketten D. 1984. Correlations of morphology with frequency for
Odontocete cochlea: systematics and topology. Baltimore: The
Johns Hopkins University.
Ketten DR. 1992. The marine mammal ear: specializations for
aquatic audition and echolocation. In: Webster D, Fay RR, Popper
AN, editors. The evolutionary biology of hearing. New York:
Springer-Verlag. p 717–754.
Ketten DR. 1994. Functional analyses of whale ears: adaptations
for underwater hearing. I.E.E.E. Proc Underwater Acoust 1:264–
Ketten DR. 2000. Cetacean ears. In: Fay RR, editor. Hearing by
whales and dolphins. New York: Springer-Verlag. p 43–108.
Ketten DR, Skinner MW, Wang G, Vannier MW, Gates GA, Neely
JG. 1998. In vivo measures of cochlear length and insertion depth
of nucleus cochlear implant electrode arrays. Ann Otolol Rhinol
Ketten DR, Wartzok D. 1990. Three-dimensional reconstructions of
the dolphin ear. In: Kastelein R, editor. Sensory abilities of ceta-
ceans. New York: Plenum Press. p 81–105.
Knowlton AR, Kraus SD. 2001. Mortality and serious injury of
northern right whales (Eubalaena glacialis) in the western North
Atlantic. J Cetacean Res Manage (Special Issue) 2:193–208.
Kraus SD. 1990. Rates and potential causes of mortality in North
Atlantic right whales (Eubalaena glacialis). Mar Mammal Sci
Laist DW, Knowlton AR, Mead JG, Collet AS, Podesta M. 2001. Col-
lisions between ships and whales. Mar Mammal Sci 17:35–75.
Manley GA. 1971. Some aspects of the evolution of hearing in verte-
brates. Nature 230:506–509.
Moore MJ, Knowlton AR, Kraus SD, McLellan WA, Bonde RK.
2004. Morphometry, gross morphology and available histopathol-
ogy in North Atlantic right whale (Eubalaena glacialis) mortal-
ities (1970 to 2002). J Cetacean Res Manage 6:199–214.
Nadol JB. 1988. Quantification of human spiral ganglion cells by se-
rial section reconstruction and segmental density estimates. Am J
Nowacek DP, Johnson MP, Tyack PL. 2003. North Atlantic right
whales (Eubalaena glacialis) ignore ships but respond to alerting
stimuli. Proc Biol Sci 271:227–231.
Nummela S, Thewissen JGM, Bajpai S, Hussain ST, Kumar K.
2007. Sound transmission in archaic and modern whales: anatom-
ical adaptations for underwater hearing. Anat Rec (this issue).
Parks SE. 2003. Response of North Atlantic right whales (Euba-
laena glacialis) to playback of calls recorded from surface active
groups in both the North and South Atlantic. Mar Mammal Sci
Parks SE, Tyack PL. 2005. Sound production by North Atlantic
right whales (Eubalaena glacialis) in surface active groups. J
Acoust Soc Am 117:3297–3306.
Richardson WJ, Greene CR Jr, Malme CI, Thomson DH. 1995. Ma-
rine mammals and noise. San Diego: Academic Press.
Sales GD, Pye D. 1974. Ultrasonic communication by animals. Lon-
don: Chapman and Hall.
Schuknecht HF. 1993. Pathology of the ear. 2 ed. Philadelphia: Lea
Tyack P. 1983. Differential response of humpback whales, Mega-
ptera novaeangliae, to playback of song or social sounds. Behav
Ecol Sociobiol 13:49–55.
Urick RJ. 1983. Principles of underwater sound. Los Altos, CA: Pen-
Wartzok D, Ketten DR. 1999. Marine mammal sensory systems. In:
Reynolds JE III, Rommel SA, editors. Biology of marine mam-
mals. Washington: Smithsonian Institution Press. p 117–175.
Watkins WA. 1981. Activities and underwater sounds of fin whales.
Sci Rep Whales Res Inst 33:83–117.
Watkins WA. 1986. Whale reactions to human activities in Cape
Cod waters. Mar Mammal Sci 2:251–262.
West CD. 1985. The relationship of the spiral turns of the cochlea
and the length of the basilar membrane to the range of audible
frequencies in ground dwelling mammals. J Acoust Soc Am
PARKS ET AL.