Modeling whale audiograms: effects of bone mass on high-frequency hearing.

In a previous paper (Hemilä et al., Hear. Res. 133 (1999) 82-97) we have presented a mechanical model, based on species-specific anatomical data, for the toothed whale middle ear. For five odontocete species of six we found that the model quite well predicted published behavioral audiograms. Here we report that new published data indicate that the audiogram of the sixth and deviating species, the killer whale Orcinus orca, was from a specimen with deficient high-frequency hearing. A new published killer whale audiogram is similar to other odontocete audiograms and does fit our four-bone model. With certain general conditions, a model with isometric (middle) ears results in uniform audiograms for different species, when presented in a log-log plot; with larger ears the audiogram curves are just moved towards lower frequencies. The audiograms coincide in case all frequencies are scaled by a factor 1/m3, where m is the mass of the ear ossicles. Odontocete ears are isometric enough to show that the corresponding audiograms are indeed similar after such mass scaling. Specifically, this scaling factor can be used to predict the high-frequency hearing limits of all odontocete species. Our anatomical data and models support the notion that ossicular mass is a crucial factor limiting high-frequency hearing in both terrestrial mammals and toothed whales.

The anatomy of the killer whale middle ear (Orcinus orca).

The paper first reviews our present understanding of the functional morphology of the odontocete (toothed whale) ear. The tympano-periotic complex forming the ear region consists of a ventral bowl-shaped tympanic bone in direct contact with the surrounding soft tissues and the incident sound, and a dorsal periotic bone containing the inner ear. Apparently sound brings the tympanic bone, and especially its thin tympanic plate, into vibration. The ossicles in the air-filled middle ear cavity form a bridge from the tympanic plate to the periotic bone connecting the vibrating plate to the oval window and the inner ear. Our computer tomography (CT) sections and camera lucida drawings reveal two hitherto unknown features of the odontocete ear, both of them of potential relevance to sound reception and impedance matching. (1) It is well known that, in addition to the ossicular chain, two other bone structures connect the tympanic to the periotic bone. We show that the most delicate parts of these extra-ossicular connections consist of thin and folded bony sheets which apparently allow compliance in the tympano-periotic bone contacts and enable plate vibration in relation to the periotic bone. (2) The round head of the malleus, in combination with a fitting round depression on the periotic side, seems to form a joint. We propose that this (hypothetical) joint, together with the adjacent structures, forms a lever producing an amplification of the vibration velocity at the level of the oval window.

Scaling of the cetacean middle ear.

Functionally interesting dimensions of the tympano-periotic complex were measured and compared in 18 odontocete and six mysticete species, ranging from small porpoises to the blue whale. We determined (i) the masses of the tympanic and periotic bones (T and P) and of the ossicles malleus, incus, and stapes (M, I and S), (ii) the volume occupied bythe tympanic bone (V), (iii) the areas of the tympanic plate and oval window (A1 and A2), (iv) the thickness of the tympanic plate (D), and (v) the densities of the ossicles (dM, dI, and dS). In most cases, roughly isometric scaling was found in both toothed and baleen whales. P is isometric to T, and the tympanic bone is structurally isometric in all species studied, although not within mysticetes as a group, shown by the isometric relations of V to T, of T(2/3) to A1, and of D to square root(A1). The essentially isometric scaling of the tympanic bone provides a basis for the functional models described by Hemilä et al. (1999). The relation of S to M+I is also isometric, but the relation of M+I+S to T is negatively allometric, as is the relation of A2 to A1, both with slopes close to 2/3. The possible functional implication of this allometry is unknown. The mean ossicular density is 2.64 g/cm3 for odontocetes, and 2.35 g/cm3 for mysticetes. The highly mineralized and convex tympanic plate provides cetaceans with a uniquely large and stiff sound collecting area.

A model of the odontocete middle ear.

The high acoustic sensitivity of the bottlenose dolphin is physically defined and related to the anatomy of the middle ear. The paper presents a conceptual and parametric analysis of the demands imposed by this high sensitivity upon the middle ear mechanisms: the head and the middle ear structures must collect sound energy from a large area and concentrate it onto the oval window. Assuming that the specific input impedance of the mammalian cochlea is relatively constant, and smaller than the characteristic acoustic impedance of water, we find that the impedance matching task of the cetacean middle ear is very different from that of terrestrial mammals: instead of a large pressure amplification, cetaceans need amplification of particle velocity. Our mechanical four-bone model of the odontocete middle ear is based on the anatomy of the tympano-periotic complex and consists of four rigid bone units (tympanic bone, the malleus-incus complex, stapes, periotic bone) connected through elastic junctions. The velocity amplification is brought about by lever mechanisms and elastic couplings. The model produced velocity amplifications ranging from 7- to 23-fold when provided with middle ear parameters from the six odontocete species for which audiograms are available. The model reproduces the complete audiograms of these six species fairly well for frequencies up to about 100-120 kHz.

Noise-equivalent and signal-equivalent visual summation of quantal events in space and time.

Noise recorded in visual neurons, or variability in psychophysical experiments, may be quantified in terms of quantal fluctuations from an "equivalent" steady illumination. The conversion requires assumptions concerning how photon signals are pooled in space and time, i.e. how to pass from light fluxes to numbers of photon events relevant to the Poisson statistics describing signal/noise. It is usual to approximate real weighting profiles for the integration of photon events in space and time (the sensitivity distribution of the receptive field [RF] and the waveform of the impulse response [IR]) by sharp-bordered apertures of "complete," equal-weight summation of events. Apertures based on signal-equivalence cannot provide noise-equivalence, however, because greater numbers of events summed with smaller weights (as in reality) have lower variances than smaller numbers summed with full weight. Thus sharp-bordered apertures are necessarily smaller if defined for noise- than for signal-equivalence. We here consider the difference for some commonly encountered RF and IR profiles. Summation areas, expressed as numbers of photoreceptors (cones or rods) contributing with equal weight, are denoted NS for signal and NN for noise; sharply delimited summation times are correspondingly denoted tS and tN. We show that the relation in space is NN = 0.5 NS for the Gaussian distribution (e.g. for the RF center mechanism of retinal ganglion cells). For a photoreceptor in an electrically coupled network the difference is even larger, e.g., for rods in the toad retina NN = 0.2 NS (NS = 13.7 rods and NN = 2.8 rods). In time, the relation is tN approximately 0.7 tS for realistic quantal response waveforms of photoreceptors. The surround input in a difference-of-Gaussians RF may either decrease or increase total noise, depending on the degree of correlation of center and surround noise. We introduce a third useful definition of sharp-bordered summation apertures: one that provides the same signal-to-noise ratio (SNR) for large-long stimuli as the real integration profiles. The SNR-equivalent summation area is N* = N2S/NN and summation time t* = t2S/tN.

Light adaptation of cone photoresponses studied at the photoreceptor and ganglion cell levels in the frog retina.

The sensitivity and time scale of the dominant (562 nm) cone system of the frog, Rana temporaria, were studied as functions of steady adapting illuminance (IB). Photoreceptor responses to brief flashes of light were recorded as aspartate-isolated ERG mass potentials from the isolated retina. The characteristics of the cone signal after transmission through the retina were derived from response thresholds and stimulus--intensity-response--latency functions for extracellularly recorded spike discharges of single ganglion cells in the eyecup. At 14 degrees C, the single-photon response of dark-adapted cones, extrapolated from ERG intensity-response functions, had an amplitude of 0.5% of the saturated response (Umax) and peaked at tp approximately 0.4 sec. Steady background illumination decreased both tp and flash sensitivity (SF), starting from apparent "dark lights" of, respectively, less than 10 (for time scale) and about 100 (for sensitivity) photoisomerisations per cone per second [P*sec-1]. From there upwards, two distinct ranges of background adaptation were apparent. Under moderate backgrounds (up to IB approximately 10(4) - 10(5) P*sec-1), sensitivity fell according to the relation SF alpha IB-0.64 and time scale shortened according to tp alpha IB-0.16. Under brighter backgrounds, from approx. 10(5) P*sec-1 up to the limit of our light source at 10(7) P*sec-1, the decrease in SF was significantly stronger than predicted by the Weber relation (SF alpha IB-1), while the decrease in tp levelled out and even tended to reverse. All these changes were virtually identical at the photoreceptor and ganglion cell levels, although the absolute time scale of cone signals apparent at the latter level was 2-fold longer. Our general conclusion is that photoreceptors have several distinct regimes for light adaptation, and traditional descriptions of functional changes (in sensitivity and kinetics) relevant to vision need to be restated with higher resolution, in view also of recent insights into the diversity of underlying mechanisms.

pH changes in frog rods upon manipulation of putative pH-regulating transport mechanisms.

Rod intracellular pH (pHi) in the intact frog retina was measured fluorometrically with the dye 2',7'-bis(2-carboxyethyl)-5(and-6)-carboxyfluorescein under treatments chosen to affect putative pH-regulating transport mechanisms in the plasma membrane. The purpose was to relate possible pHi changes to previously reported effects on photoresponses. In nominally bicarbonate-free Ringer, application of amiloride (1 mM) or substitution of 95 mM external Na+ by K+ or choline triggered monotonic but reversible acidifications, consistent with inhibition of Na+/H+ exchange. Bicarbonate-dependent mechanisms were characterized as follows: (1) Replacing half of a 12 mM phosphate buffer by bicarbonate caused a sustained rise of pHi. (2) Subsequent application of the anion transport inhibitor 4,4'-diisothiocyanatostilbene-2',2'-disulphonic acid (DIDS, 0.2 mM) set off a slow acidification. (3) Substitution of external Cl- by gluconate (95 mM) caused a rapid pHi rise both in normal Na+ and low-Na+ perfusion. (4) This effect was inhibited by DIDS. The results support a consistent explanation of parallel electrophysiological experiments on the assumption that intracellular acidifications reduce and alkalinizations (in a certain range) augment photoresponses. It is concluded that both Na+/H+ exchange and bicarbonate transport control rod pHi, modulating the light-sensitive current. Part of the bicarbonate transport is by Na(+)-independent HCO3-/Cl- exchange, but a further Na(+)-coupled bicarbonate import mechanism is implicated.

Modelling the spatio-temporal modulation response of ganglion cells with difference-of-Gaussians receptive fields: relation to photoreceptor response kinetics.

Difference-of-Gaussians (DOG) models for the receptive fields of retinal ganglion cells accurately predict linear responses to both periodic stimuli (typically moving sinusoidal gratings) and aperiodic stimuli (typically circular fields presented as square-wave pulses). While the relation of spatial organization to retinal anatomy has received considerable attention, temporal characteristics have been only loosely connected to retinal physiology. Here we integrate realistic photoreceptor response waveforms into the DOG model to clarify how far a single set of physiological parameters predict temporal aspects of linear responses to both periodic and aperiodic stimuli. Traditional filter-cascade models provide a useful first-order approximation of the single-photon response in photoreceptors. The absolute time scale of these, plus a time for retinal transmission, here construed as a fixed delay, are obtained from flash/step data. Using these values, we find that the DOG model predicts the main features of both the amplitude and phase response of linear cat ganglion cells to sinusoidal flicker. Where the simplest model formulation fails, it serves to reveal additional mechanisms. Unforeseen facts are the attenuation of low temporal frequencies even in pure center-type responses, and the phase advance of the response relative to the stimulus at low frequencies. Neither can be explained by any experimentally documented cone response waveform, but both would be explained by signal differentiation, e.g. in the retinal transmission pathway, as demonstrated at least in turtle retina.

Changes in retinal time scale under background light: observations on rods and ganglion cells in the frog retina.

The kinetics of rod responses to flashes and steps of light was studied as a function of background intensity (IB) at the photoreceptor and ganglion cell levels in the frog retina. Responses of the rod photoreceptors were recorded intracellularly in the eyecup and as ERG mass potentials across the isolated, aspartate-superfused retina. The kinetics of the retinally transmitted signal was derived from the latencies of ganglion cell spike discharges recorded extracellularly in the eyecup. In all states of adaptation the linear-range rod response to dim flashes could be modelled as the impulse response of a chain of low-pass filters with the same number of stages: 4 (ERG) or 4-6 (intracellular). Dark-adapted time-to-peak (tp, mean +/- SD) at 12 degrees C was 2.4 +/- 0.6 sec (ERG) or 1.7 +/- 0.4 sec (intracellular). Under background light, the time scale shortened as a power function of background intensity, I-bB with b = 0.19 +/- 0.03 (ERG) or 0.14 +/- 0.04 (intracellular). The latency-derived time scale of the rod-driven signal at the ganglion cell agreed well with that of the photoreceptor responses. The apparent underlying impulse response had tp = 2.0 +/- 0.7 sec in darkness and accelerated as I-bB with b = 0.17 +/- 0.03. The photoreceptor-to-ganglion-cell transmission delay shortened by 30% between darkness and a background delivering ca 10(4) photoisomerizations per rod per second. Data from the literature suggest that all vertebrate photoreceptors may accelerate according to similar power functions of adapting intensity, with exponents in the range 0.1-0.2. It is noteworthy that the time scale of human (foveal) vision in experiments on flicker sensitivity and temporal summation shortens as a power function of mean luminance with b approximately 0.15.

What middle ear parameters tell about impedance matching and high frequency hearing.

Acoustic energy enters the mammalian cochlea aided by an anatomical impedance matching performed by the middle ear. The purpose of this paper is to analyse the functional consequences of changes in scale of the middle ear when going from the smallest mammals to the largest. Our anatomical measurements in mammals of different sizes ranging from bats to elephants indicate that middle ear proportions are largely isometric. Thus the calculated transformer ratio is basically independent of animal size, a typical value lying between 30 and 80. Similarly, the calculated specific acoustic input impedance of the inner ear is independent of animal size, the average value being about 140 kPa s/m. We show that if the high frequency hearing limit of isometric ears is limited by ossicle inertia, it should be inversely proportional to the cubic root of the ossicular mass. This prediction is in reasonable agreement with published audiogram data. We then present a three-parameter model of the middle ear where some obvious deviations from perfect isometry are taken into account. The high frequency hearing limits of different species generally agree well with the predictions of this simple model. However, the hearing limits of small rodents clearly deviate from the model calculation. We interpret this observation as indicating that the hearing limit towards very high frequencies may be set by cochlear transduction mechanisms. Further we discuss the exceptional high frequency hearing of the cat and the amphibious hearing of seals.

Spectral sensitivities of short- and long-wavelength sensitive cone mechanisms in the frog retina.

ERG mass photoreceptor responses were recorded across the isolated, aspartate-perfused retina of the frog, Rana temporaria, in order to determine spectral sensitivities of cones. Cone responses were distinguished from rod responses by their faster kinetics, and responses from different cone types were isolated by selective background adaptation. Our main finding is that of a novel short-wavelength sensitive cone population peaking at about 431 nm. Further, we find that the sensitivity spectrum of the dominant long-wavelength sensitive cone population fully accounts for the most common type of photopic ganglion cell spectrum. Both can be described by a nomogram with lambda max = 562 nm. This resolves a long-standing apparent conflict between cone absorbance spectra and ganglion cell sensitivities. Including the 502 nm cones previously described by microspectrophotometry, the frog possesses a collection of cones that could support trichromatic photopic vision.

Changes in the light-sensitive current of salamander rods upon manipulation of putative pH-regulating mechanisms in the inner and outer segment.

The light-sensitive current of dark-adapted rods isolated from the Ambystoma retina was recorded while either the inner or the outer segment (IS or OS) protruding from the suction pipette was exposed to treatments intended to reveal the physiological roles of pH-regulating transport mechanisms. Applied to the IS, both amiloride (presumed to block Na+/H+ exchange, 2 mM) and 4-4'-diisothiocyanatostilbene-2,2'-disulphonic acid (DIDS) (presumed to block bicarbonate transport, 0.1 mM) generally abolished light sensitivity completely but reversibly, consistent with acidification of the IS. Yet, the circulating ("dark") current often persisted, implying that the OS was not acidified. Applied to the OS, amiloride depressed but DIDS increased the dark current and photoresponses. Given the fact that the current increases with rising OS-pHi, this suggests alkalinization, which could be due to DIDS inhibiting bicarbonate extrusion by HCO3-/Cl- exchangers in the OS. Consistent with this idea, replacing external Cl- by other anions increased the current as would be expected if HCO3-/Cl- exchange is reversed. We propose that the IS and OS manage their acid balances independently and with different sets of transport mechanisms. Acidosis in either compartment suppresses the photosensitivity of the rod, but by differing mechanisms.

pH regulation in frog cones studied by mass receptor photoresponses from the isolated retina.

Mass cone photoresponses were recorded across the aspartate-treated frog retina under treatments chosen to affect putative pH-regulating mechanisms. The saturated response amplitude (Umax) was found to be a monotonically increasing function of perfusion pH in the range 7-8, and thus presumably of intracellular pH (pHi). Accepting that Umax can be used as an index of pHi changes, two results indicate the importance of bicarbonate transport for preventing intracellular acidification: (1) bicarbonate-buffered (6 mM HCO3- + 6 mM HEPES) perfusate increased Umax compared with nominally bicarbonate-free perfusate (12 mM HEPES); (2) the anion transport blocker DIDS (0.1 mM) caused a strong decrease in the amplitude of photoresponses. Substitution of 95 mM chloride by gluconate in the perfusing fluid boosted photoresponses indicating that at least part of the bicarbonate transport involves HCO3-/Cl- exchange. Amiloride (2 mM) also caused a decrease of photoresponse amplitude, which suggests that Na+/H+ exchange contributes to pHi regulation. In all these respects, cones behaved similarly to rods. Cones differed from rods (in the intact retina) in that addition of 0.5 mM of the carbonic anhydrase inhibitor acetazolamide reduced (never augmented) photoresponses. The difference is considered in relation to the presence of carbonic anhydrase in cone, as opposed to rod, outer segments.

On the relation between ERG waves and retinal function: inverted rod photoresponses from the frog retina.

In rod mass receptor photoresponses recorded across the isolated frog retina, a paradoxical cornea-positive wave may precede the response of normal polarity. We present a model which shows that the light-induced decrease in rod current can give rise to inverted or biphasic ERG signals if the distal part (tip) of the rod outer segment responds more slowly and/or less sensitively than the proximal part (base). The condition is that current entering at the tip is represented with greater weight in the ERG. The model reproduces recorded ERG waveforms well. It further predicts that if there is a light-insensitive conductance in the tip membrane, ERG photoresponses may be non-recordable although current photoresponses are only slightly reduced. The model reveals a type of complexity in the relation between mass potentials and underlying physiological processes which has not previously received attention.

Rod phototransduction modulated by bicarbonate in the frog retina: roles of carbonic anhydrase and bicarbonate exchange.

1. Effects on rod phototransduction following manipulation of retinal CO2-HCO3- and H+ fluxes were studied in dark-adapted retinas of the frog and the tiger salamander. 2. Rod photoresponses to brief flashes of light were recorded from the isolated sensory retina as electroretinogram mass receptor potentials and from isolated rods by the suction-pipette technique. The experimental treatments were: (1) varying [CO2] + [HCO3-] in the perfusion fluid: (2) applying acetazolamide (AAA), which inhibits the enzyme carbonic anhydrase (CA); and (3) applying 4,4'-diisothiocyanatostilbene-2,2'-disulphonic acid (DIDS) which blocks exchange mechanisms transporting HCO3- across cell membranes. 3. The concentration of the internal transmitter of the rods, cyclic GMP, was biochemically determined from the rod outer segment layer of retinas that had been incubated in the same solutions as were used for perfusion in the electrophysiological experiments. 4. The introduction of 6 mM-sodium bicarbonate to replace half the buffer of a nominally CO2-HCO3(-)-free (12 mM-phosphate or HEPES, [Na+] constant) Ringer solution doubled the cyclic GMP concentration in the rod outer segment layer and increased the saturating response amplitude and the relative sensitivity of rods in the intact retina. 5. The introduction of 0.5 mM-AAA into bicarbonate-containing Ringer solution accelerated the growth of saturated responses and sensitivity. Incubation of the retina in AAA-bicarbonate Ringer solution elevated the concentration of cyclic GMP ninefold compared with the phosphate control. 6. No effects of switching to bicarbonate-AAA Ringer solution were observed in the photocurrent of isolated rods drawn into suction pipettes with only the outer segment protruding into the perfusion fluid. The target of AAA is probably the CA-containing Müller cell. 7. The introduction of DIDS into the perfusate (at normal pH 7.5) set off a continuous decay of photoresponses which finally abolished light sensitivity completely. The decay proceeded regardless of whether bicarbonate and AAA were present or not. 8. Rods that had lost their photosensitivity in DIDS recovered almost fully when the pH of the DIDS perfusate was raised to 8.5. They also recovered when DIDS was washed out with bicarbonate Ringer solution at constant pH (7.5). 9. It is proposed that all our treatments ultimately modulate the intracellular pH of the rods which is determined by the relative rates of H+ leakage and HCO3- transport into the cells.(ABSTRACT TRUNCATED AT 400 WORDS)

Sulfhydryl binding reagents increase the conductivity of the light-sensitive channel and inhibit phototransduction in retinal rods.

The mechanisms by which sulfhydryl (SH-) binding reagents modulate the light-sensitive conductance of retinal rods were investigated by current recording from single rods, by patch clamp recording from the plasma membrane of the rod outer segment (ROS), and by biochemical study of their effects on the light-induced hydrolysis of cyclic GMP. The electrophysiology, as well as measurements of the reagents' ability to traverse the ROS plasma membrane, was done on amphibian (Rana and Ambystoma) rods, and the biochemistry on bovine rods. The main SH-reagents used were N-ethyl-maleimide (NEM) and iodoacetamide (IAA). Both transiently increased rod current, but part of the large current could not be turned off by light. After a few minutes' exposure, NEM, but not IAA, caused a continuous decay of the rod's light sensitivity. In patch-clamp recordings from the ROS plasma membrane, the reagents increased conductivity both in the presence and absence of cGMP, consistent with the observation that the drug-induced current increase in intact rods involved both light-sensitive and light-insensitive components. In vitro, NEM was found to be a powerful inhibitor of cGMP hydrolysis, which can explain the gradual loss of light sensitivity in the rod and could initially contribute to the increased dark current via elevated cGMP levels. Thus, SH-reagents act both by modifying the light-sensitive channel and by inhibiting phototransduction inside the rod.

Signal transmission through the dark-adapted retina of the toad (Bufo marinus). Gain, convergence, and signal/noise.

Responses to light were recorded from rods, horizontal cells, and ganglion cells in dark-adapted toad eyecups. Sensitivity was defined as response amplitude per isomerization per rod for dim flashes covering the excitatory receptive field centers. Both sensitivity and spatial summation were found to increase by one order of magnitude between rods and horizontal cells, and by two orders of magnitude between rods and ganglion cells. Recordings from two hyperpolarizing bipolar cells showed a 20 times response increase between rods and bipolars. At absolute threshold for ganglion cells (Copenhagen, D.R., K. Donner, and T. Reuter. 1987. J. Physiol. 393:667-680) the dim flashes produce 10-50-microV responses in the rods. The cumulative gain exhibited at each subsequent synaptic transfer from the rods to the ganglion cells serves to boost these small amplitude signals to the level required for initiation of action potentials in the ganglion cells. The convergence of rod signals through increasing spatial summation serves to decrease the variation of responses to dim flashes, thereby increasing the signal-to-noise ratio. Thus, at absolute threshold for ganglion cells, the convergence typically increases the maximal signal-to-noise ratio from 0.6 in rods to 4.6 in ganglion cells.

Temperature-dependence of rod photoresponses from the aspartate-treated retina of the frog (Rana temporaria).

The effects of temperature changes on rod photoresponses were studied by recording the aspartate-isolated mass receptor potential in the dark-adapted retina of the frog Rana temporaria. The amplitude of saturating responses, indicating the magnitude of the dark current, increased linearly with temperature in the measured range 6-26 degrees C, extrapolating to zero dark current at 0 degrees C. Sensitivity was maximal around 18 degrees C but the decrease towards lower temperatures was shallow. The results show that rod phototransduction in the frog Rana temporaria is adapted to lower temperatures than in the tropical toad Bufo marinus. Responses to dim flashes were, approximately up to peak, well fitted by the same 'independent activation' model with four delay stages as have been found to best describe current responses from single toad rods. The kinetics (reciprocal time-to-peak) showed Arrhenius-type temperature-dependence with apparent activation energy 12.4 kcal mol-1 and Q10 = 2.1.

Transient sensitivity reduction and biphasic photoresponses observed when frog retinal rods are oxidized.

1. Rod photoresponses and the effects of oxidation have been studied by recording either the transretinal voltage in aspartate-treated retinas or the outer segment current of single rods. 2. Oxidizing conditions transiently decreased, reducing conditions increased sensitivity. 3. Biphasic photoresponses were seen when the level of oxidation was rising and also in some other sensitivity-depressing conditions. 4. A model is proposed which explains the biphasic responses in terms of sensitivity differences between the tip and the base of the rod outer segment.