Mammalian Hearing Mechanisms: Functional Modeling

20

Mammalian Hearing Mechanisms: Functional Modeling

Hearing capacities are the output of the integrated components of the whole ear. All

mammalian ears, including those of marine mammals, have three basic divisions:

1) an outer ear,

2) an air-filled middle ear with bony levers and membranes, and

3) a fluid-filled inner ear with mechanical resonators and sensory cells.

The outer ear acts as a sound collector. The middle ear transforms acoustic components into mechanical ones detectable by the inner ear. The inner ear acts as a band-pass filter and mechano-chemical transducer of sound into neural impulses.

Outer and Middle Ears

The outer ear is subdivided conventionally into a pinna or ear flap that assists in localization,

a funnel-shaped concha, and the ear canal or auditory tube. The size and shape of each

component in each species is extraordinarily diverse, which makes any generalized statement

about the function of the outer ear debatable. In most mammals, the pinnal flaps are distinct

flanges that may be mobile. These flanges act as sound diffractors that aid in localization,

primarily by acting as a funnel that selectively admits sounds along the pinnal axis (Heffner and

Heffner 1992).

The middle ear is commonly described as an impedance-matching device or transformer that

counteracts the ~36 dB loss from the impedance differences between air and the fluid-filled inner

ear, an auditory hangover of vertebrate movement from water onto land. This gain is achieved

by the mechanical advantages provided by the difference in the area of the middle ear membranes

(large tympanic vs. small oval window) and by the lever ratio of the bony chain of middle ear

ossicles which creates a pressure gain and a reduction in particle velocity at the inner ear.

Improving the efficiency of power transfer to the inner ear may not, however, be the only

function for the middle ear. Recent studies on land mammals have led to a competing (but not

mutually exclusive) theory called the peripheral filter-isopower function, in which the middle ear

has a "tuning" role (see Zwislocki 1981, Rosowski 1994, Yost 1994 for comprehensive

discussions). The middle ear is an air-filled cavity with significant differences among species in

volume, stiffness (K), and mass (M). Each species has a characteristic middle ear resonance

based on the combined chain of impedances, which, in turn, depends upon the mechanical

properties of its middle ear components. For any animal, the sum of impedances is lowest; i. e.,

21

middle ear admittance is greatest and energy transmission most efficient, at the middle ear's

resonant frequency (f). As expected, this frequency also tends to be at or near the frequency

with the lowest threshold (best sensitivity) for that species (Fay 1992).

Stiffness and mass have inverse effects on frequency in a resonant system:

f = (1/2p) (8)

Put another way, mass dominated systems have a lower resonant frequency than stiffness

dominated systems. Increasing stiffness in any ear component (membranes, ossicles, cavity)

improves the efficiency of transmission of high frequencies. Adding mass to the system, e.g., by

increasing cavity volume or increasing ossicular chain mass, favors low frequencies.

Consequently, in addition to impedance matching, middle ears may be evolutionarily tuned as

evidenced by different combinations of mass or stiffening agents in each species. Ultrasonic

species like microchiropteran bats and dolphins have ossicular chains stiffened with bony struts

and fused articulations (Reysenbach de Haan 1956, Pye 1972, Sales and Pye 1974, Ketten and

Wartzok 1990). Low frequency species, like heteromyid desert rodents, mole rats, elephants,

and mysticetes, have large, middle ears with flaccid tympanic membranes (Webster 1962;

Hinchcliffe and Pye 1969; Webster and Webster 1975; Fleischer 1978; Ketten 1992, 1994).

Inner Ear

Mammalian inner ears are precocial; i.e., they are structurally mature and functional at birth

and may be active in utero. Inner ears are similarly tuned in that inner ear stiffness and mass

characteristics are major determinants of species-specific hearing ranges. The inner ear consists

of the cochlea (primary hearing receptor) and the vestibular system (organs of orientation and

balance) (Fig. 4).

The cochlea is a fluid-filled spiral with a resonator, the basilar membrane, and a

neuroreceptor, the Organ of Corti (Figure 5). When the basilar membrane moves, cilia on the

hair cells of the Organ of Corti are deflected eliciting chemical changes that release

neurotransmitters. Afferent fibers of the auditory nerve synapsing on the hair cells carry acoustic

details to the brain, including frequency, amplitude, and temporal patterning, based on the

location, degree of deflection, and sequencing of hair cells that are excited by basilar membrane

motion. Efferent fibers also synapse with the hair cells, but their function is not yet fully

understood. As discussed in the final sections, damage the hair cells is the primary mechanism

underlying most hearing loss.

A key component in the cochlear system is the basilar membrane. Differences in hearing

ranges are dictated largely by differences in stiffness and mass of the basilar membrane that are

the result of basilar membrane thickness and width variations along the cochlear spiral. From

base (closest the oval and round windows) to apex (farthest from the middle ear), changes in the

construction of the basilar membrane in each mammal mechanically tune the ear to a specific set

of frequencies (Figure 4). Each membrane region has a particular resonance characteristic and

consequently greater deflection than other regions of the membrane for some input frequency.

For any input signal within the hearing range of the animal, the entire basilar membrane will

22

respond to some degree. At any one moment, each region of the membrane will have a different

amount of deflection and a different phase related to the input signal. Over time, changes in

amplitude and phase at each point give the impression of a traveling response wave along the

cochlea, but because the membrane segments that have resonance characteristics closest to

frequencies in the signal have greater displacements than other segments of the membrane, a

characteristic profile or envelope develops for the signal. Figure 4 shows the place-dependent

differences in the displacement envelopes that would occur in a generic mammalian inner ear for

three pure-tone inputs.

Basilar membrane dimensions vary inversely, and generally regularly, with cochlear

dimensions. The highest frequency each animal hears is encoded at the base of the cochlear

spiral (near the oval window), where the membrane is narrow, thick, and stiff. Moving towards

the apex of the spiral, as the membrane becomes broader and more pliant, progressively lower

frequencies are encoded. Therefore, mammalian basilar membranes are essentially tonotopically

arranged resonator arrays, ranging high to low from base to apex, rather like a guitar with

densely packed strings graded to cover multiple octaves.

Recall that, in general, small mammals have good high frequency hearing characteristics and

large mammals have comparatively low hearing ranges. Early inner ear models were based on

the assumption that all mammalian basilar membranes were constructed of similar components

that had a constant gradient with length and that length scaled with animal size. On average,

smaller animals were assumed to have shorter, narrower, stiffer membranes while larger animals

had longer membranes in which the majority of membrane modules were broader and less stiff

(von Békésy 1960; Greenwood 1961, 1990). Given that assumption, frequency distributions in

the inner ear of any species could be derived by comparing one parameter, basilar membrane

length, with an arbitrary standard, the average human membrane length. For many land

mammals, this assumption is correct, but only because length is an indirect correlate of other key

features for basilar membrane resonance. For these ears, now termed "generalists" (Fay 1992;

Echteler et al. 1994), basilar membrane thickness and width covary regularly with length;

therefore, length can proportionately represent stiffness.

Only recently has it become clear that some species, termed "specialists" (Echteler et al.

1994), do not have the same thickness-width-length relationship as generalist land mammals

(Manley 1972, Ketten 1984, 1997; West 1986). Most specialist animals have retuned their inner

ears to fit an atypical tuning for their body size by either increasing mass to improve low

frequency sensitivity in small ears (as in mole rats) or adding stiffening components to increase

resonant frequencies in larger inner ears (as in dolphins) (Hinchcliffe and Pye 1969; Sales and

Pye 1974; Webster and Webster 1975; Ketten 1984). The most extreme case of specialization is

to be found in some bats which have relatively constant basilar membrane dimensions for ~30%

of the cochlea and thereby devote a disproportionate amount of the membrane to encoding a

very narrow band frequencies related to a component of their echolocation signal (Bruns and

Schmieszek 1980, Vater 1988a, Kössl and Vater 1995).

Structure-function-habitat links

Marine mammal ears fall into both categories and some species have a mix of generalist and

specialist traits. Like land mammals, pinnipeds and cetaceans have basilar membranes that scale

with animal size. Consequently, because marine mammals are relatively large, most have basilar

membranes longer than the human average. If marine mammal ears followed the generalist land

mammal pattern, most would have relatively poor ultrasonic hearing. For example, standard land

mammal length-derived hearing models (Greenwood 1961, 1990; Fay 1992) predict an upper

limit of hearing of ~16 kHz for bottlenosed dolphins, Tursiops truncatus, which actually have a

functional high frequency hearing limit of 160 kHz (Au 1993). Prior to the discovery of dolphin

echolocation, it was assumed that these large animals had predominately low functional hearing

ranges similar to cows. Hearing is not constrained to low frequencies in marine mammals

because they have radically different inner ear thickness-width gradients than generalist land

mammals. In odontocetes, very high ultrasonic hearing is related also to the presence of

extensive stiffening additions to the inner ear. These features, discussed in detail later in the

document, demonstrate the usefulness of comparative audiometric and anatomical studies for

teasing apart hearing mechanisms. In fact, one important outgrowth of marine mammal hearing

studies has been the development of multi-feature hearing models that are better predictors of

hearing characteristics for all mammals than traditional, single-dimension models (Ketten, 1994,

1997).