THE ANATOMICAL RECORD 297:892–900 (2014)
Elastic Modulus of Cetacean Auditory Ossicles ANDREW A. TUBELLI,1* ALEKS ZOSULS,1 DARLENE R. KETTEN,2,3 1 AND DAVID C. MOUNTAIN 1 Department of Biomedical Engineering, Hearing Research Center, Boston University, Boston, Massachusetts 2 Department of Biology, Woods Hole Oceanographic Institution, Marine Research Facility, Woods Hole, Massachusetts 3 Department of Otology and Laryngology, Harvard Medical School, Massachusetts Eye and Ear Infirmary, Boston, Massachusetts
ABSTRACT In order to model the hearing capabilities of marine mammals (cetaceans), it is necessary to understand the mechanical properties, such as elastic modulus, of the middle ear bones in these species. Biologically realistic models can be used to investigate the biomechanics of hearing in cetaceans, much of which is currently unknown. In the present study, the elastic moduli of the auditory ossicles (malleus, incus, and stapes) of eight species of cetacean, two baleen whales (mysticete) and six toothed whales (odontocete), were measured using nanoindentation. The two groups of mysticete ossicles overall had lower average elastic moduli (35.2 6 13.3 GPa and 31.6 6 6.5 GPa) than the groups of odontocete ossicles (53.3 6 7.2 GPa to 62.3 6 4.7 GPa). Interior bone generally had a higher modulus than cortical bone by up to 36%. The effects of freezing and formalin-fixation on elastic modulus were also investigated, although samples were few and no clear trend could be discerned. The high elastic modulus of the ossicles and the differences in the elastic moduli between mysticetes and odontocetes are likely specializations in the bone for underwater hearing. Anat Rec, 297:892– C 2014 Wiley Periodicals, Inc. 900, 2014. V
Key words: cetacean; ossicles; middle ear; hearing; elastic modulus
INTRODUCTION Information on the mechanical properties of bone in the marine mammal (cetacean) auditory system is limited. Cetaceans possess an auditory system that is anatomically similar to that of terrestrial mammals but is morphologically different, being highly specialized for underwater use (Ketten, 2000). Cetacean ossicles are more massive than their terrestrial counterparts; additionally, the anterior process of the malleus forms a synostosis with the tympanic bone, a characteristic also found in few terrestrial mammals such as bats, shrews, and mice (Fleischer, 1978). In vivo measurement of hearing capability in cetaceans is restricted by marine mammal protection laws, and behavioral and electrophysiological methods of obtaining hearing information tell us little about the mechanics of the auditory system. A way around C 2014 WILEY PERIODICALS, INC. V
this issue is to simulate sound reception with biologically realistic models (e.g. Gan et al., 2004; Elkhouri et al., 2006; Cranford et al., 2008; Homma et al., 2009; Tubelli et al., 2012).
Grant sponsor: Joint Industry Program and the Living Marine Resources program; Grant number: N00244-100-0053. *Corresponding to: Andrew A. Tubelli, Hearing Research Center and Department of Biomedical Engineering, Boston University, 44 Cummington St., Boston, MA 02215. Fax: 617-3536766. E-mail: [email protected] Received 12 January 2013; Revised 9 September 2013; Accepted 15 January 2014. DOI 10.1002/ar.22896 Published online 13 February 2014 in Wiley Online Library (wileyonlinelibrary.com).
893
ELASTIC MODULUS CETACEAN AUDITORY OSSICLES
Transmission of sound through the cetacean middle ear is not well understood. Biophysical models provide the best evidence for how sound can travel through the middle ear. These models are highly dependent on the mechanical property values of tissues that are used as parameters. Elastic modulus, a key mechanical property that is used to describe the stiffness of a material, is one such mechanical property. Elastic modulus is defined as the ratio of stress per unit area that a material undergoes divided by the resultant strain (deformation). In order to calculate the frequency response of cetaceans as accurately as possible (Tubelli et al., 2012), the most realistic values of elastic modulus must be used. There are few publications that describe the mechanical properties of the ear bones, especially of cetacean ear bones. The research groups that have measured the elastic modulus of cetacean ear bones are summarized in Table 1. Each study used a different method to measure elastic modulus and each obtained significantly different results. There are only a couple known studies that measure the material properties of terrestrial mammalian ossicles, in human (Speirs et al., 1999) and rabbit (Soons et al., 2010). These values are also listed in Table 1. The values obtained for these two terrestrial species are considerably less than those of cetaceans. This difference in ossicular stiffness between terrestrial and aquatic mammals, when combined with the anatomical similarities and morphological differences, suggests a similar yet modified mechanism for sound reception in cetaceans. The ossicles of the middle ear are the three smallest bones found in all mammals. The ossicles are also irregularly shaped. Figure 1 shows anatomical reconstructions of the intact ossicular chain in a mysticete (baleen whale) and an odontocete (toothed whale) species. As a result, conventional mechanical tests such as bending and compression tests, which require regularly shaped samples of a certain size, are not suitable to measure material properties, such as elastic modulus, of the ossicles. The tympanic bones are much larger than the ossicles, thereby allowing bending tests to be used in Currey (1979) and Zioupos et al. (1997). Speirs et al.
(1999) used compression testing on the ossicles but noted the artifact associated with using such small samples. Lees et al. (1996) used sonic velocity and density measurements to calculate elastic moduli with the caveat that the results were to be regarded as representative of elasticity rather than regarded as precise values. Instead, nanoindentation is a more suitable method of elastic modulus measurement. Nanoindentation uses a specialized tip that indents a material of interest and can determine material properties at the level of the microstructure. Nanoindentation has been used to measure the elastic modulus in a number of studies of other bones (e.g. Zysset et al., 1999; Fan and Rho, 2003; Donnelly et al., 2005). Additionally, since bone is heterogeneous with a complex structure, nanoindentation can measure the modulus of different regions of the bone rather than treating the modulus as a bulk parameter. The goal of this study was to measure the elastic moduli of the ossicles via nanoindentation for eight cetacean species: minke whale (Balaenoptera acutorostrata), fin whale (Balaenoptera physalus), common dolphin (Delphinus delphis), short-finned pilot whale (Globicephala macrorhynchus), Atlantic white-sided dolphin (Lagenorhynchus acutus), harbor porpoise (Phocoena phocoena), striped dolphin (Stenella coeruleoalba), and bottlenose dolphin (Tursiops truncatus). The sensitivity of these properties to dehydration and to fixation by freezing or formalin was also examined.
MATERIALS AND METHODS Sample Preparation A total of 45 ossicles were obtained from 17 cetacean specimens under a permit for scientific research on cetacean tissues issued to D. Ketten. Table 2 lists all specimens used. Due to the nature of some of the deaths, not all information could be collected. Specimens ranged from calf to adult. Full maturation of the cetacean tympanic bulla occurs 1 year after birth (de Buffrenil et al., 2004), so most animals were presumed to have fully developed ossicles, with the exception of the bottlenose
TABLE 1. Average value of elastic modulus (E) measured for auditory bones Source Cetacean Bulla Currey (1979) Lees et al. (1996)
Bone B. physalus tympanic bulla B. physalus tympanic bulla B. physalus malleus B. physalus periotic T. truncatus periotic
Zioupos et al. (2000) Zioupos (2005) Terrestrial ossicles Speirs et al. (1999) Soons et al. (2010)
B. physalus tympanic bulla B. physalus tympanic bulla Human malleus/incus Rabbit malleus/incus Caput malleus Collum malleus Corpus incudis Crus longum incudis
Average E 6 standard deviation (GPa)
Method Three-point bending Estimation from sonic and density Estimation from sonic and density Estimation from sonic and density Estimation from sonic and density Three-point bending Nanoindentation
velocity
31.3 6 0.97 51.9
velocity
56.5
velocity
52
velocity
73.6
Compression Two-needle indentation
33.5 40.2 6 5.1 2–3 16.4 6 2.8 16.3 6 2.9 15.6 6 1.8 16.8 6 3.1 17.1 6 3.8
894
TUBELLI ET AL.
dolphin calf. This animal may or may not have had fully developed ossicles. There were no known pathologies associated with the auditory bones in any of the specimens. Four ears were preserved via formalin fixation and one ear was preserved via freezing. The approximate duration of time that each ear was preserved is indicated in Table 1. Within the specified duration, the ears may have been thawed and refrozen or taken out of and placed back into formalin. This would have been within a 48-h timeframe for either observation or nondestructive experimentation. The remaining non-preserved ears were chilled at 4 C without preservation and disarticulated within several days after animal death. The bones were cleaned of soft tissue with dental tools and then dehydrated in air. The bones were then embedded in underwater epoxy (Sea Goin’ Poxy Putty, Permalite Plastics Inc., Costa Mesa, CA) onto a 15 mm magnetic specimen disc, cured in an oven at 38 C for 1 h and then left at room temperature to cure for at least 24 h more. Heating is not expected to have any effect on the bone samples. Collagen in bone does not start to denature until 60 C; furthermore, collagen denaturation has little effect on elastic modulus up to 200 C (Wang et al., 2001). Once the epoxy cured, the samples were ground flat using a miniature milling machine (Micro-Mark, Berkeley Heights, NJ). Samples were kept hydrated with distilled water during milling. The resulting surface was polished using a sequence of silicon carbide abrasive paper with progressively finer grit sizes (400, 600, 1,500) and finished with 0.5 and 0.1 micron diamond paste on low nap polishing cloth. The surface was washed off with distilled water and blotted dry. Atomic force microscopy was used to verify that the polishing technique produced a smooth enough surface. The resulting RMS roughness for a polished specimen was on the order of 2 to 23 nm.
Nanoindentation The indentation was done using a Triboindenter (Hysitron Inc., Minneapolis, MN). The system uses the Oliver-Pharr method (Oliver and Pharr, 1992) to determine the reduced modulus, the effective modulus of both the indenter and the specimen together, via the unloading portion of a force-displacement curve. The reduced modulus is calculated using Eq. (1), where S is the contact stiffness, calculated as the slope of a power-law fit over a region of 95% to 20% of the unloading forcedisplacement curve, and Ac is the contact area.
pffiffiffi p S pffiffiffiffiffiffi Er 5 2 AC
Fig. 1. Anatomical reconstructions of (a) a minke whale and (b) a bottlenose dolphin ossicular chain, medial view. Orange: segment of tympanic bone, red: malleus, green: incus, blue: stapes. ap: anterior process, mb: manubrial region of the body of the malleus, cb: central region of the body of the malleus, mh: head/articular region of the malleus, bo: body of the incus, lp: long process of the incus, sp: short process of the incus, he: head of the stapes, ba: base of the stapes, ac: anterior crus, pc: posterior crus. The short process of the minke whale incus is occluded. The black scale bar in each panel corresponds to 5 mm in length.
(1)
The elastic modulus of the sample can then be calculated from the reduced modulus using Eq. (2), where the Poisson’s ratio and the elastic modulus of the tip (mtip and Etip, respectively) are known. A diamond Berkovich tip was used (approximately 150 nm radius) for indentation, with mtip equal to 0.07 and Etip equal to 1,140 GPa. 2
2
1 ð12vsample Þ ð12vtip Þ 5 1 Etip Er Esample
(2)
The Poisson’s ratio of the sample (msample) was assumed to be 0.3, a value commonly used for bone, for
M
B-acu 23
Unknown Unknown
T-tru 53 T-tru 75 T-tru 198
Bottlenose dolphin
Ossicles are abbreviated: malleus (M), incus (I), stapes (S). Note that both left and right ears were available for D-del 60.
F
S-coe 08
Striped dolphin
F
F
P-pho 99
F
D-del 61
M
M
D-del 60
L-acu 46
Unknown
D-del 59
F
Globicephala macrorhynchus Lagenorhynchus acutus Phocoena phocoena Stenella coeruleoalba Tursiops truncatus
M
B-acu 22
G-mac 64
Common dolphin
Delphinus delphis
F
B-acu 18
Unknown
F
B-acu 17
B-phy 12
M
Sex
B-acu 15
Specimen
Short-finned pilot whale Atlantic white-sided dolphin Harbor porpoise
Fin whale
Balaenoptera physalus
Common name
Minke whale
Balaenoptera acutorostrata
Species name
Adult
Unknown
Calf
Adult
Adult
Adult
Adult
Adult
Adult
Sub-adult
Unknown
Sub-adult
Sub-adult
Adult
Juvenile
Juvenile
Age
Stranded in tidal pool, died at site, fresh tissue Stranded on beach, died at site, fresh tissue Stranded on beach, found dead, fresh tissue Bycatch in fisheries net, died at site, fresh tissue Found stranded, fresh tissue Mass stranding, died at site, fresh tissue Mass stranding, died at site, fresh tissue Stranded on beach, died of old age, fresh tissue
Stranded on beach, euthanized, fresh tissue Stranded on rock jetty, died at site, fresh tissue Found floating with line marks from entanglement, moderately decomposed Found dead and entangled in water, moderately decomposed Stranded on beach, found dead, moderately decomposed Found several days after death, moderately decomposed Bycatch in fisheries net, died at site, fresh tissue Bycatch in fisheries net, died at site, fresh tissue
Condition found
TABLE 2. Specimen information Ear
I,S
Right
Right
Left
Left
Left
Right
Right
M,I,S
M,I
M,I,S
M,I,S
M,I,S
M,I,S
M,I,S
M,I,S M,I,S
Right Left Left
M,I,S
M,I,S
I,S
Left
Right
Left
M,I,S
I,S
Left
Right
M,I,S
I
Ossicles
Right
Left
Fixation
None
Formalin, 58 months
Formalin, 109 months
None
None
None
None
None None
None
None
None
None
None
Frozen, 16 days
Formalin, 35–39 months Formalin, 57 months
ELASTIC MODULUS CETACEAN AUDITORY OSSICLES
895
896
TUBELLI ET AL.
this analysis. Varying the value of Poisson’s ratio between 0.2 and 0.4 has little effect on the elastic modulus (Rho et al., 1997). Nanoindentation was performed at various regions on the flat polished surface of the bone samples at room temperature. All indents were load-controlled and used a load/unload rate of 600 mN/s. A maximum load of 6,000 mN was used to provide sufficient depth penetration and thereby prevent surface roughness from affecting the data (Donnelly et al., 2005); indentation depth ranged from approximately 300 to 500 nm. A hold time of 30 s was used to account for the viscoelastic properties of bone (Fan and Rho, 2003). Calibration of the data was done using a quartz sample (reduced modulus of 69.9 GPa). Regions were chosen based on anatomical location. These locations are shown in Fig. 2 with the exception of the short process of the incus. More detailed information on the macrostructure of cetacean ossicles can be found in sources such as Fraser and Purves (1960), Fleischer (1978), Solntseva (2007), and Mead and Fordyce (2009). Since the ossicles were oriented randomly within the epoxy, indents were made only in regions that could be seen on the final polished surface. Two categorizations of bone were made for ossicular measurements: an outer (“cortical”) layer and an inner, less uniformly calcified lacunate core (“interior”). In humans, the ossicles are composed of a mixture of enchondral and intrachondrial bone with a dense outer cortex and a less dense interior characterized by scattered lacunae (Gulya, 2007). Thus, the distribution of these bone types in cetacean ossicles was found to be morphologically similar to that observed in human ossicles (Kirikae, 1960; Galioto and Marley, 1965). These distinct bone regions (here referred to as "cortex" and "interior") were clearly distinguishable in the histological sections and are labeled in Fig. 2. It is important to note that the predominant bony elements of the ossicles, which include particularly intrachondrial bone, are quite unlike those of long bones, such as the femur, which are typically used for bone mechanical property measurements. Indentation of the cortical region was not always possible because the cortex can be very thin in some portions. Indentation near the boneepoxy boundary was avoided. Indentation regions were typically defined as a square area of 0.64 mm2. A total of 16 indents were performed in each indentation region. Spacing between individual indents was not uniform, but all indents were spaced greater than 5 lm apart. In the majority of indentations performed, there was a single indentation region per anatomical region;
Fig. 2. Histological slices of (a) a striped dolphin malleus, (b) a common dolphin incus, and (c) a harbor porpoise stapes. The slices are representative of the surface of a nanoindentation sample. Histology was completed by first decalcifying the sample in EDTA and then embedding the sample in celloidin. The sample was sectioned at 25 lm, and every 10th section was stained with hematoxylin and eosin and mounted on glass slides. Bone types and anatomical regions distinguished in experiments are labeled, with the exception the short process of the incus (not shown). ap: anterior process, mb: manubrial region of the body of the malleus, cb: central region of the body of the malleus, mh: head/articular region of the malleus, bo: body of the incus, lp: long process of the incus, he: head of the stapes, ba: base of the stapes, ac: anterior crus, pc: posterior crus. The black scale bar in each panel corresponds to 1 mm in length.
ELASTIC MODULUS CETACEAN AUDITORY OSSICLES
however, there were four cases in which there were two indentation regions per anatomical region. The interior manubrial region of a bottlenose dolphin malleus (T-tru 53, left ear) was indented twice to test variability within an anatomical region. The cortical region of the long process of the incus was indented twice in three species (Bacu 22, right ear; B-acu 23, right ear; G-mac 64, left ear). On the cut surface of the bone of these three samples, the thick cortex could be seen flanking the interior bone. One indentation region was assigned to the cortex on each side of the interior bone, again to test variability within an anatomical region. All samples were kept dry until indented. The samples were rehydrated in 0.9% saline a half hour before indentation. Rehydration of bone after drying was found to have an insignificant effect on elasticity (Currey, 1988).
RESULTS There were a few samples where some data points were excluded from analysis. For the Atlantic whitesided dolphin, the long process of the incus exhibited several large cracks from sample preparation that affected the integrity of the region. For the bottlenose dolphin, suspected machine error occurred on the same day as indentation of four regions, three incudal and one mallear. All indents from each of these regions were ignored during analysis. There were four specimens that were formalin-fixed before indentation, two minke whale ears and two bottlenose dolphin ears. Both formalin fixed minke ears had ossicles that had a higher elastic modulus than their fresh counterparts: malleus, 49% higher; incus, 32% higher; stapes, 12% higher. For the bottlenose dolphin, however, the malleus and incus moduli were 6% and 33% lower, respectively, and the stapes modulus was 11% higher. One specimen, a minke whale ear, was fro-
897
zen and thawed before indentation. The thawed incus was 14% higher than its fresh counterpart and the thawed stapes was 11% lower than its fresh counterpart. Since the sample size was so low and variability among members of the same species is uncharacterized, no clear trend can be discerned for preserved bone. Both interior and cortical regions were indented in 21 ossicles, five of which were fixed. Formalin and frozen ossicles did not show any difference in trend from the fresh ossicles. In 15 of the 21 ossicles, interior bone overall had a higher elastic modulus than cortical bone (Fig. 3). In most of these ossicles, the average modulus of interior bone was 1% to 36% higher than the average modulus of cortical bone. The fin whale was a special case, where interior bone was twice as stiff as cortical bone. Looking at cortical vs. interior within specific regions, there are 15 cases of a statistically significant difference between cortical and interior bone (unpaired t-test, P
Elastic Modulus of Cetacean Auditory Ossicles ANDREW A. TUBELLI,1* ALEKS ZOSULS,1 DARLENE R. KETTEN,2,3 1 AND DAVID C. MOUNTAIN 1 Department of Biomedical Engineering, Hearing Research Center, Boston University, Boston, Massachusetts 2 Department of Biology, Woods Hole Oceanographic Institution, Marine Research Facility, Woods Hole, Massachusetts 3 Department of Otology and Laryngology, Harvard Medical School, Massachusetts Eye and Ear Infirmary, Boston, Massachusetts
ABSTRACT In order to model the hearing capabilities of marine mammals (cetaceans), it is necessary to understand the mechanical properties, such as elastic modulus, of the middle ear bones in these species. Biologically realistic models can be used to investigate the biomechanics of hearing in cetaceans, much of which is currently unknown. In the present study, the elastic moduli of the auditory ossicles (malleus, incus, and stapes) of eight species of cetacean, two baleen whales (mysticete) and six toothed whales (odontocete), were measured using nanoindentation. The two groups of mysticete ossicles overall had lower average elastic moduli (35.2 6 13.3 GPa and 31.6 6 6.5 GPa) than the groups of odontocete ossicles (53.3 6 7.2 GPa to 62.3 6 4.7 GPa). Interior bone generally had a higher modulus than cortical bone by up to 36%. The effects of freezing and formalin-fixation on elastic modulus were also investigated, although samples were few and no clear trend could be discerned. The high elastic modulus of the ossicles and the differences in the elastic moduli between mysticetes and odontocetes are likely specializations in the bone for underwater hearing. Anat Rec, 297:892– C 2014 Wiley Periodicals, Inc. 900, 2014. V
Key words: cetacean; ossicles; middle ear; hearing; elastic modulus
INTRODUCTION Information on the mechanical properties of bone in the marine mammal (cetacean) auditory system is limited. Cetaceans possess an auditory system that is anatomically similar to that of terrestrial mammals but is morphologically different, being highly specialized for underwater use (Ketten, 2000). Cetacean ossicles are more massive than their terrestrial counterparts; additionally, the anterior process of the malleus forms a synostosis with the tympanic bone, a characteristic also found in few terrestrial mammals such as bats, shrews, and mice (Fleischer, 1978). In vivo measurement of hearing capability in cetaceans is restricted by marine mammal protection laws, and behavioral and electrophysiological methods of obtaining hearing information tell us little about the mechanics of the auditory system. A way around C 2014 WILEY PERIODICALS, INC. V
this issue is to simulate sound reception with biologically realistic models (e.g. Gan et al., 2004; Elkhouri et al., 2006; Cranford et al., 2008; Homma et al., 2009; Tubelli et al., 2012).
Grant sponsor: Joint Industry Program and the Living Marine Resources program; Grant number: N00244-100-0053. *Corresponding to: Andrew A. Tubelli, Hearing Research Center and Department of Biomedical Engineering, Boston University, 44 Cummington St., Boston, MA 02215. Fax: 617-3536766. E-mail: [email protected] Received 12 January 2013; Revised 9 September 2013; Accepted 15 January 2014. DOI 10.1002/ar.22896 Published online 13 February 2014 in Wiley Online Library (wileyonlinelibrary.com).
893
ELASTIC MODULUS CETACEAN AUDITORY OSSICLES
Transmission of sound through the cetacean middle ear is not well understood. Biophysical models provide the best evidence for how sound can travel through the middle ear. These models are highly dependent on the mechanical property values of tissues that are used as parameters. Elastic modulus, a key mechanical property that is used to describe the stiffness of a material, is one such mechanical property. Elastic modulus is defined as the ratio of stress per unit area that a material undergoes divided by the resultant strain (deformation). In order to calculate the frequency response of cetaceans as accurately as possible (Tubelli et al., 2012), the most realistic values of elastic modulus must be used. There are few publications that describe the mechanical properties of the ear bones, especially of cetacean ear bones. The research groups that have measured the elastic modulus of cetacean ear bones are summarized in Table 1. Each study used a different method to measure elastic modulus and each obtained significantly different results. There are only a couple known studies that measure the material properties of terrestrial mammalian ossicles, in human (Speirs et al., 1999) and rabbit (Soons et al., 2010). These values are also listed in Table 1. The values obtained for these two terrestrial species are considerably less than those of cetaceans. This difference in ossicular stiffness between terrestrial and aquatic mammals, when combined with the anatomical similarities and morphological differences, suggests a similar yet modified mechanism for sound reception in cetaceans. The ossicles of the middle ear are the three smallest bones found in all mammals. The ossicles are also irregularly shaped. Figure 1 shows anatomical reconstructions of the intact ossicular chain in a mysticete (baleen whale) and an odontocete (toothed whale) species. As a result, conventional mechanical tests such as bending and compression tests, which require regularly shaped samples of a certain size, are not suitable to measure material properties, such as elastic modulus, of the ossicles. The tympanic bones are much larger than the ossicles, thereby allowing bending tests to be used in Currey (1979) and Zioupos et al. (1997). Speirs et al.
(1999) used compression testing on the ossicles but noted the artifact associated with using such small samples. Lees et al. (1996) used sonic velocity and density measurements to calculate elastic moduli with the caveat that the results were to be regarded as representative of elasticity rather than regarded as precise values. Instead, nanoindentation is a more suitable method of elastic modulus measurement. Nanoindentation uses a specialized tip that indents a material of interest and can determine material properties at the level of the microstructure. Nanoindentation has been used to measure the elastic modulus in a number of studies of other bones (e.g. Zysset et al., 1999; Fan and Rho, 2003; Donnelly et al., 2005). Additionally, since bone is heterogeneous with a complex structure, nanoindentation can measure the modulus of different regions of the bone rather than treating the modulus as a bulk parameter. The goal of this study was to measure the elastic moduli of the ossicles via nanoindentation for eight cetacean species: minke whale (Balaenoptera acutorostrata), fin whale (Balaenoptera physalus), common dolphin (Delphinus delphis), short-finned pilot whale (Globicephala macrorhynchus), Atlantic white-sided dolphin (Lagenorhynchus acutus), harbor porpoise (Phocoena phocoena), striped dolphin (Stenella coeruleoalba), and bottlenose dolphin (Tursiops truncatus). The sensitivity of these properties to dehydration and to fixation by freezing or formalin was also examined.
MATERIALS AND METHODS Sample Preparation A total of 45 ossicles were obtained from 17 cetacean specimens under a permit for scientific research on cetacean tissues issued to D. Ketten. Table 2 lists all specimens used. Due to the nature of some of the deaths, not all information could be collected. Specimens ranged from calf to adult. Full maturation of the cetacean tympanic bulla occurs 1 year after birth (de Buffrenil et al., 2004), so most animals were presumed to have fully developed ossicles, with the exception of the bottlenose
TABLE 1. Average value of elastic modulus (E) measured for auditory bones Source Cetacean Bulla Currey (1979) Lees et al. (1996)
Bone B. physalus tympanic bulla B. physalus tympanic bulla B. physalus malleus B. physalus periotic T. truncatus periotic
Zioupos et al. (2000) Zioupos (2005) Terrestrial ossicles Speirs et al. (1999) Soons et al. (2010)
B. physalus tympanic bulla B. physalus tympanic bulla Human malleus/incus Rabbit malleus/incus Caput malleus Collum malleus Corpus incudis Crus longum incudis
Average E 6 standard deviation (GPa)
Method Three-point bending Estimation from sonic and density Estimation from sonic and density Estimation from sonic and density Estimation from sonic and density Three-point bending Nanoindentation
velocity
31.3 6 0.97 51.9
velocity
56.5
velocity
52
velocity
73.6
Compression Two-needle indentation
33.5 40.2 6 5.1 2–3 16.4 6 2.8 16.3 6 2.9 15.6 6 1.8 16.8 6 3.1 17.1 6 3.8
894
TUBELLI ET AL.
dolphin calf. This animal may or may not have had fully developed ossicles. There were no known pathologies associated with the auditory bones in any of the specimens. Four ears were preserved via formalin fixation and one ear was preserved via freezing. The approximate duration of time that each ear was preserved is indicated in Table 1. Within the specified duration, the ears may have been thawed and refrozen or taken out of and placed back into formalin. This would have been within a 48-h timeframe for either observation or nondestructive experimentation. The remaining non-preserved ears were chilled at 4 C without preservation and disarticulated within several days after animal death. The bones were cleaned of soft tissue with dental tools and then dehydrated in air. The bones were then embedded in underwater epoxy (Sea Goin’ Poxy Putty, Permalite Plastics Inc., Costa Mesa, CA) onto a 15 mm magnetic specimen disc, cured in an oven at 38 C for 1 h and then left at room temperature to cure for at least 24 h more. Heating is not expected to have any effect on the bone samples. Collagen in bone does not start to denature until 60 C; furthermore, collagen denaturation has little effect on elastic modulus up to 200 C (Wang et al., 2001). Once the epoxy cured, the samples were ground flat using a miniature milling machine (Micro-Mark, Berkeley Heights, NJ). Samples were kept hydrated with distilled water during milling. The resulting surface was polished using a sequence of silicon carbide abrasive paper with progressively finer grit sizes (400, 600, 1,500) and finished with 0.5 and 0.1 micron diamond paste on low nap polishing cloth. The surface was washed off with distilled water and blotted dry. Atomic force microscopy was used to verify that the polishing technique produced a smooth enough surface. The resulting RMS roughness for a polished specimen was on the order of 2 to 23 nm.
Nanoindentation The indentation was done using a Triboindenter (Hysitron Inc., Minneapolis, MN). The system uses the Oliver-Pharr method (Oliver and Pharr, 1992) to determine the reduced modulus, the effective modulus of both the indenter and the specimen together, via the unloading portion of a force-displacement curve. The reduced modulus is calculated using Eq. (1), where S is the contact stiffness, calculated as the slope of a power-law fit over a region of 95% to 20% of the unloading forcedisplacement curve, and Ac is the contact area.
pffiffiffi p S pffiffiffiffiffiffi Er 5 2 AC
Fig. 1. Anatomical reconstructions of (a) a minke whale and (b) a bottlenose dolphin ossicular chain, medial view. Orange: segment of tympanic bone, red: malleus, green: incus, blue: stapes. ap: anterior process, mb: manubrial region of the body of the malleus, cb: central region of the body of the malleus, mh: head/articular region of the malleus, bo: body of the incus, lp: long process of the incus, sp: short process of the incus, he: head of the stapes, ba: base of the stapes, ac: anterior crus, pc: posterior crus. The short process of the minke whale incus is occluded. The black scale bar in each panel corresponds to 5 mm in length.
(1)
The elastic modulus of the sample can then be calculated from the reduced modulus using Eq. (2), where the Poisson’s ratio and the elastic modulus of the tip (mtip and Etip, respectively) are known. A diamond Berkovich tip was used (approximately 150 nm radius) for indentation, with mtip equal to 0.07 and Etip equal to 1,140 GPa. 2
2
1 ð12vsample Þ ð12vtip Þ 5 1 Etip Er Esample
(2)
The Poisson’s ratio of the sample (msample) was assumed to be 0.3, a value commonly used for bone, for
M
B-acu 23
Unknown Unknown
T-tru 53 T-tru 75 T-tru 198
Bottlenose dolphin
Ossicles are abbreviated: malleus (M), incus (I), stapes (S). Note that both left and right ears were available for D-del 60.
F
S-coe 08
Striped dolphin
F
F
P-pho 99
F
D-del 61
M
M
D-del 60
L-acu 46
Unknown
D-del 59
F
Globicephala macrorhynchus Lagenorhynchus acutus Phocoena phocoena Stenella coeruleoalba Tursiops truncatus
M
B-acu 22
G-mac 64
Common dolphin
Delphinus delphis
F
B-acu 18
Unknown
F
B-acu 17
B-phy 12
M
Sex
B-acu 15
Specimen
Short-finned pilot whale Atlantic white-sided dolphin Harbor porpoise
Fin whale
Balaenoptera physalus
Common name
Minke whale
Balaenoptera acutorostrata
Species name
Adult
Unknown
Calf
Adult
Adult
Adult
Adult
Adult
Adult
Sub-adult
Unknown
Sub-adult
Sub-adult
Adult
Juvenile
Juvenile
Age
Stranded in tidal pool, died at site, fresh tissue Stranded on beach, died at site, fresh tissue Stranded on beach, found dead, fresh tissue Bycatch in fisheries net, died at site, fresh tissue Found stranded, fresh tissue Mass stranding, died at site, fresh tissue Mass stranding, died at site, fresh tissue Stranded on beach, died of old age, fresh tissue
Stranded on beach, euthanized, fresh tissue Stranded on rock jetty, died at site, fresh tissue Found floating with line marks from entanglement, moderately decomposed Found dead and entangled in water, moderately decomposed Stranded on beach, found dead, moderately decomposed Found several days after death, moderately decomposed Bycatch in fisheries net, died at site, fresh tissue Bycatch in fisheries net, died at site, fresh tissue
Condition found
TABLE 2. Specimen information Ear
I,S
Right
Right
Left
Left
Left
Right
Right
M,I,S
M,I
M,I,S
M,I,S
M,I,S
M,I,S
M,I,S
M,I,S M,I,S
Right Left Left
M,I,S
M,I,S
I,S
Left
Right
Left
M,I,S
I,S
Left
Right
M,I,S
I
Ossicles
Right
Left
Fixation
None
Formalin, 58 months
Formalin, 109 months
None
None
None
None
None None
None
None
None
None
None
Frozen, 16 days
Formalin, 35–39 months Formalin, 57 months
ELASTIC MODULUS CETACEAN AUDITORY OSSICLES
895
896
TUBELLI ET AL.
this analysis. Varying the value of Poisson’s ratio between 0.2 and 0.4 has little effect on the elastic modulus (Rho et al., 1997). Nanoindentation was performed at various regions on the flat polished surface of the bone samples at room temperature. All indents were load-controlled and used a load/unload rate of 600 mN/s. A maximum load of 6,000 mN was used to provide sufficient depth penetration and thereby prevent surface roughness from affecting the data (Donnelly et al., 2005); indentation depth ranged from approximately 300 to 500 nm. A hold time of 30 s was used to account for the viscoelastic properties of bone (Fan and Rho, 2003). Calibration of the data was done using a quartz sample (reduced modulus of 69.9 GPa). Regions were chosen based on anatomical location. These locations are shown in Fig. 2 with the exception of the short process of the incus. More detailed information on the macrostructure of cetacean ossicles can be found in sources such as Fraser and Purves (1960), Fleischer (1978), Solntseva (2007), and Mead and Fordyce (2009). Since the ossicles were oriented randomly within the epoxy, indents were made only in regions that could be seen on the final polished surface. Two categorizations of bone were made for ossicular measurements: an outer (“cortical”) layer and an inner, less uniformly calcified lacunate core (“interior”). In humans, the ossicles are composed of a mixture of enchondral and intrachondrial bone with a dense outer cortex and a less dense interior characterized by scattered lacunae (Gulya, 2007). Thus, the distribution of these bone types in cetacean ossicles was found to be morphologically similar to that observed in human ossicles (Kirikae, 1960; Galioto and Marley, 1965). These distinct bone regions (here referred to as "cortex" and "interior") were clearly distinguishable in the histological sections and are labeled in Fig. 2. It is important to note that the predominant bony elements of the ossicles, which include particularly intrachondrial bone, are quite unlike those of long bones, such as the femur, which are typically used for bone mechanical property measurements. Indentation of the cortical region was not always possible because the cortex can be very thin in some portions. Indentation near the boneepoxy boundary was avoided. Indentation regions were typically defined as a square area of 0.64 mm2. A total of 16 indents were performed in each indentation region. Spacing between individual indents was not uniform, but all indents were spaced greater than 5 lm apart. In the majority of indentations performed, there was a single indentation region per anatomical region;
Fig. 2. Histological slices of (a) a striped dolphin malleus, (b) a common dolphin incus, and (c) a harbor porpoise stapes. The slices are representative of the surface of a nanoindentation sample. Histology was completed by first decalcifying the sample in EDTA and then embedding the sample in celloidin. The sample was sectioned at 25 lm, and every 10th section was stained with hematoxylin and eosin and mounted on glass slides. Bone types and anatomical regions distinguished in experiments are labeled, with the exception the short process of the incus (not shown). ap: anterior process, mb: manubrial region of the body of the malleus, cb: central region of the body of the malleus, mh: head/articular region of the malleus, bo: body of the incus, lp: long process of the incus, he: head of the stapes, ba: base of the stapes, ac: anterior crus, pc: posterior crus. The black scale bar in each panel corresponds to 1 mm in length.
ELASTIC MODULUS CETACEAN AUDITORY OSSICLES
however, there were four cases in which there were two indentation regions per anatomical region. The interior manubrial region of a bottlenose dolphin malleus (T-tru 53, left ear) was indented twice to test variability within an anatomical region. The cortical region of the long process of the incus was indented twice in three species (Bacu 22, right ear; B-acu 23, right ear; G-mac 64, left ear). On the cut surface of the bone of these three samples, the thick cortex could be seen flanking the interior bone. One indentation region was assigned to the cortex on each side of the interior bone, again to test variability within an anatomical region. All samples were kept dry until indented. The samples were rehydrated in 0.9% saline a half hour before indentation. Rehydration of bone after drying was found to have an insignificant effect on elasticity (Currey, 1988).
RESULTS There were a few samples where some data points were excluded from analysis. For the Atlantic whitesided dolphin, the long process of the incus exhibited several large cracks from sample preparation that affected the integrity of the region. For the bottlenose dolphin, suspected machine error occurred on the same day as indentation of four regions, three incudal and one mallear. All indents from each of these regions were ignored during analysis. There were four specimens that were formalin-fixed before indentation, two minke whale ears and two bottlenose dolphin ears. Both formalin fixed minke ears had ossicles that had a higher elastic modulus than their fresh counterparts: malleus, 49% higher; incus, 32% higher; stapes, 12% higher. For the bottlenose dolphin, however, the malleus and incus moduli were 6% and 33% lower, respectively, and the stapes modulus was 11% higher. One specimen, a minke whale ear, was fro-
897
zen and thawed before indentation. The thawed incus was 14% higher than its fresh counterpart and the thawed stapes was 11% lower than its fresh counterpart. Since the sample size was so low and variability among members of the same species is uncharacterized, no clear trend can be discerned for preserved bone. Both interior and cortical regions were indented in 21 ossicles, five of which were fixed. Formalin and frozen ossicles did not show any difference in trend from the fresh ossicles. In 15 of the 21 ossicles, interior bone overall had a higher elastic modulus than cortical bone (Fig. 3). In most of these ossicles, the average modulus of interior bone was 1% to 36% higher than the average modulus of cortical bone. The fin whale was a special case, where interior bone was twice as stiff as cortical bone. Looking at cortical vs. interior within specific regions, there are 15 cases of a statistically significant difference between cortical and interior bone (unpaired t-test, P
