The authors have declared that no competing interests exist.
Conceived and designed the experiments: TSM FL. Performed the experiments: TSM BM. Analyzed the data: TSM FL. Wrote the paper: TSM BM FL.
Several teleost species have evolved anterior extensions of the swim bladder which come close to or directly contact the inner ears. A few comparative studies have shown that these morphological specializations may enhance hearing abilities. This study investigates the diversity of swim bladder morphology in four Asian and African cichlid species and analyzes how this diversity affects their hearing sensitivity.
We studied swim bladder morphology by dissections and by making 3D reconstructions from high-resolution microCT scans. The auditory sensitivity was determined in terms of sound pressure levels (SPL) and particle acceleration levels (PAL) using the auditory evoked potential (AEP) recording technique. The swim bladders in
Our results indicate that anterior swim bladder extensions seem to improve mean absolute auditory sensitivities by 21–42 dB (SPLs) and 21–36 dB (PALs) between 0.5 and 1 kHz. Besides anterior extensions, the size of the swim bladder appears to be an important factor for extending the detectable frequency range (up to 3 kHz).
In fishes that possess swim bladders or other gas filled cavities in proximity to the inner ears, sound can stimulate the inner ears in two ways. In the direct stimulation pathway a sound source leads to the lagged movement of the denser otolith relative to the fish’s body and the sensory epithelium. Thus the otolith acts as an accelerometer in the inner ear that stimulates the hair cells of the sensory epithelium
Several species that belong to different teleost families and orders have evolved anterior swim bladder extensions coming close to or contacting the inner ears leading to improved hearing abilities (for an overview see
Cichlids provide a good example to investigate effects of swim bladder morphology on hearing abilities because this species-rich family displays a high diversity in their swim bladder morphology. This diversity ranges from swim bladders which are small or completely absent in some species of the genus
Our study aimed to improve the understanding of the effects of different swim bladder morphology (from small and without anterior extensions to highly specialized) on hearing abilities in terms of hearing bandwidth and auditory sensitivities. We asked the question of whether cichlid species with anterior swim bladder extensions possess higher auditory sensitivities and/or a broader hearing bandwidth than species without such extensions. We thus aimed to test the hypothesis that a close swim bladder-inner ear association results in an improved sensitivity above several hundred Hertz and/or the ability of fishes to detect higher frequencies.
We chose four cichlid species displaying either (i) small swim bladders (
Swim bladder preparation | microCT and 3D reconstructions | Auditory measurements | |||||||
Species |
|
SL (mm) | Fixative |
|
SL (mm) | Fixative/Staining |
|
SL (mm) | BW (g) |
|
6 | 42–63 | 10% F (4) 70% EtOH (2) | – | – | – | 8 | 42–65 | 1.0–4.6 |
|
4 | 42–55 | 10% F (3) 70% EtOH (1) | – | – | – | 7 | 44–57 | 2.3–5.2 |
|
5 | 47–75 | 10% F (4) 70% EtOH (1) | 3 | 47–54 | 10% F I2KI | 5 | 47–86 | 3.4–25.2 |
|
4 | 34–41 | 10% F (4) | 4 | 33–41 | 10% F I2KI | 8 | 33–41 | 1.3–2.8 |
EtOH, ethanol; F, formalin; I2KI, Lugol solution (2.5% potassium iodide (KI), 1.25% iodine metal (I2) in water. Numbers in parentheses indicate number of specimens subjected to the respective treatment.
Ventrolateral dissections of the swim bladder and inner ears were performed for four to six specimens (including the individuals previously used for microCT, see below) of each species. Dissecting microscopes (Wild M7 and Wild M5, Wild Heerbrugg Ltd, Heerbrugg, Switzerland) equipped with a camera lucida were used for preparations and drawings. Preparation of the individuals subjected to microCT scans showed that otoliths were not affected by the formalin fixation.
Light microscopic images were taken using a Leica M165C stereomicroscope with a DFC 290 camera, applying the multifocus option (extended focus imaging) of ImageAccess Standard 8 (Imagic AG, Glattbrugg, Switzerland).
For a detailed study of the swim bladder-inner ear relationship in
|
|
|||||||
Specimen no. | 1 | 2 | 3 | 1 | 2 | 3 | 4 | |
Fish data | SL (mm) | 47 | 54 | 49 | 36 | 40 | 41 | 33 |
BW (g) | 3.4 | 5.4 | 3.6 | 1.7 | 2.6 | 2.5 | 1.1 | |
Whole fish scan | Voxel size (µm) | x = 28 y = 28 z = 25 | x = 42 y = 42 z = 33 | – | – | x = 33 y = 33 z = 25 | – | – |
Image resolution (pixels/inch) | 900 | 600 | – | – | 770 | – | – | |
Number of images | 1799 | 1764 | – | – | 1742 | – | – | |
Close-up scan | Voxel size (µm) | – | x = 15 y = 15 z = 15 | x = 15 y = 15 z = 15 | x = 24 y = 24 z = 24 | x = 15 y = 15 z = 10 | x = 15 y = 15 z = 10 | x = 9.66 y = 9.66 z = 9.66 |
Image resolution (pixels/inch) | – | 1719 | 1719 | 1067 | 1715 | 1715 | 150 | |
Number of images | – | 981 | 980 | 495 | 996 | 996 | 497 |
BW, body weight; SL, standard length.
A SkyScan 1174 scanner employing a 50 keV/40 W tungsten X-ray source and a 1.3 megapixel CCD camera was used. The images were scanned using isotropic resolution, were reconstructed without binning, and were finally stored as BMP image stacks. A ring-artifact-reduction utility was engaged (setting 7–10) during reconstruction for all the images. For one specimen of
Prior to 3D rendering, image stacks were edited in AdobePhotoshop® CS2. Images were reduced from 16 bit to 8 bit grayscale, cropped, and, if necessary, resolution was reduced to a minimum image resolution of 600 pixels per inch for whole-fish scans and to a minimum image resolution of 1,715 pixels per inch for close-up scans (in
3D renderings of otoliths and swim bladders were performed in AMIRA® v. 5.4.0 (Visage Imaging GmbH, Berlin, Germany). A threshold-based segmentation was applied for labeling the structures and, if necessary, this labeling was refined or corrected using the brush tool. In the case of the otoliths and the swim bladder horns, every image was labeled. For the reconstruction of the posterior and middle parts of the swim bladder, initially every 10th to 20th image was labeled, with subsequent interpolation of structures on intervening images, followed by check and–if required–correction of segmentation results.
Subsequently, every otolith type as well as the swim bladder were separated from the ‘master’ LabelField file into single LabelFields and saved as separate files. The new LabelField for the swim bladder was reduced in resolution applying the Resample module. Surface rendering was performed with the SurfaceGen module. This was followed by the smoothing of surfaces using the SmoothSurface module (40 iterations for the swim bladder and 20 iterations for each ototlih type; unconstrained smoothing).
In order to visualize the in-situ position of the swim bladder and the otoliths, the whole fish was displayed using the volume rendering tool (Volren settings used: mode = maximum intensity projection, MIP; alpha = 1; color = grey.am). Finally, an overlay of the volume rendered whole fish and the 3D rendered otoliths and the swim bladder was created.
Auditory thresholds were determined by applying the auditory evoked potential (AEP) recording technique
In order to reduce muscle noise, the test subjects were immobilized with Flaxedil (gallamine triethiodide; Sigma Aldrich Handels GmbH, Vienna, Austria) at mean concentrations of 7.4, 27.2, 17.9, or 7.2 µg*g–1 body weight for
AEPs are shown 20 dB above the mean auditory threshold of each species (see
In order to make sure that AEPs were not artifacts, we tested our system with dead fishes and with no fish in the set-up. No responses were obtained from dead fishes (
In addition to sound pressure levels (SPL), we determined particle acceleration levels (PAL) at thresholds because fish species lacking hearing specializations lack sound pressure sensitivity
In order to compare SPLs and PALs for frequencies up to 1 kHz, a calibrated underwater miniature acoustic pressure-acceleration (p-a) sensor (S/N 2007-001, Applied Physical Sciences Corp., Groton, CT, USA; frequency bandwidth: 20 Hz to 2 kHz; sensitivity: −137.6 dB re 1 V/µm/s2) was placed at the fish’s position in the test tub. PALs at all stimulus frequencies and at hearing threshold levels of the fish were determined with the acceleration sensor oriented in all three orthogonal directions. In consistence with previous studies
Frequency (kHz) | SPL | PAL vert | PAL rc | PAL lat | PAL comb |
0.1 | 100 | 59 | 55 | 40 | 61 |
0.2 | 98 | 58 | 51 | 46 | 59 |
0.3 | 101 | 65 | 58 | 54 | 66 |
0.5 | 100 | 64 | 49 | 41 | 64 |
0.6 | 100 | 62 | 48 | 41 | 62 |
0.7 | 100 | 61 | 48 | 44 | 61 |
0.8 | 98 | 60 | 47 | 47 | 60 |
1.0 | 100 | 62 | 54 | 54 | 63 |
SPL-sound pressure level (dB re 1 µPa), PAL-particle acceleration level (dB re 1 µm/s2) in the vertical (vert), rostrocaudal (rc), and lateral (lat) axis; PAL comb-PAL combined of the three directions (magnitude
The spectral level of the laboratory ambient noise was measured and calculated following the methods described in
A repeated measures design was applied for analyzing potential differences between audiograms (see also
All four species possessed a swim bladder with a transverse diaphragm (including a sphincter) dividing the organ into an anterior and a posterior chamber. Except in
In the following, the inner ear-swim bladder distance is defined as the distance between the posterior most region of the ear to the anterior most part of the swim bladder. In species with extensions, this is the distance between the anterior most part of the left or right swim bladder extension to the posterior most part of the left or right ear.
The swim bladder is small and distinctly away from the inner ear. The swim bladder is shown in green; the otoliths of the inner are shown in red (lapillus = utricular otolith), pink (sagitta = saccular otolith), and yellow (asteriscus = lagenar otolith). Scale bar = 1 cm.
The swim bladder is ‘normal’ sized without contact to the inner ear. The swim bladder is shown in green; the otoliths of the inner are shown in red (lapillus), pink (sagitta), and yellow (asteriscus). Scale bar = 1 cm.
In
(
The swim bladder horn comes close to the lagena and its otolith, the asteriscus (see also
In
(
(
The air-filled part of the swim bladder is shown in shaded green and the tissue pad in dark green. The two white openings in the tissue pad (
Species |
|
|
|
|
|
|
|
|
|
8 | 7 | 5 | 8 | 8 | 7 | 5 | 8 |
Frequency (kHz) | Sound pressure level (dB re 1 µPa) | Particle acceleration level (dB re 1 µm/s2) | ||||||
0.1 | 90±1.4 | 88±1.4 | 89±2.6 | 88±1.3 | 51±1.4 | 49±1.4 | 50±2.6 | 50±1.3 |
0.2 | 84±2.4 | 87±2.7 | nm | nm | 45±2.4 | 48±2.7 | nm | nm |
0.3 | 87±2.4 | 94±2.3 | 88±3.7 | 82±1.6 | 51±2.4 | 58±2.3 | 52±3.7 | 44±1.6 |
0.5 | 103±2.3 | 112±3.5 | 82±4.2 | 70±0.8 | 66±2.3 | 75±3.5 | 45±4.2 | 39±0.8 |
0.6 | 113±1.3 | nm | nm | nm | 75±1.3 | nm | nm | nm |
0.7 | 118±1.9 | nm | nm | nm | 80±1.9 | nm | nm | nm |
0.8 | nr | 113±1.9 | 85±3.5 | 79±1.6 | nr | 75±1.9 | 47±3.5 | 40±1.6 |
1 | nr | 111±2.3 | 87±2.0 | 80±0.2 | nr | 74±2.3 | 50±2.0 | 41±0.2 |
2 | nm | 122±1.7 | 131±2.4 | 109±1.2 | ||||
3 | nm | 128±1.0 | 133±0.5 | 114±1.0 |
(
Sound pressure level (dB re 1 µPa) | Particle acceleration level (dB re 1 µm/s2) | ||||||
Source | df | MS | F | P | MS | F | P |
|
|||||||
|
|||||||
Species | 3 | 1385.123 | 31.342 |
|
1150.520 | 26.033 |
|
Error | 24 | 44.194 | |||||
|
|||||||
Frequency | 1 | 141.280 | 4,152 | 0.053 | 567.617 | 16.680 |
|
Frequency × Species | 3 | 1322.599 | 38.866 |
|
1058.075 | 31.093 |
|
Error | 24 | 34.029 | 34.029 | ||||
|
|||||||
|
|||||||
Species | 2 | 5731.730 | 178.482 |
|
5351.470 | 174.115 |
|
Error | 17 | 32.114 | 30.735 | ||||
|
|||||||
Frequency | 1 | 22468.492 | 1045.975 |
|
345.229 | 9.547 |
|
Frequency × Species | 2 | 388.819 | 18.101 |
|
1599.012 | 44.218 |
|
Error | 17 | 21.481 | 36.162 |
Significant
We tested the hypothesis that different swim bladder morphology in cichlids affects their hearing sensitivities. We found that species having swim bladder extensions directly or indirectly touching the inner ears showed a distinctly higher auditory sensitivity and/or a broader hearing bandwidth than species lacking these structures. Furthermore, we showed that the size of the swim bladder influenced hearing in addition to presence or absence of anterior swim bladder extensions.
The swim bladder morphology we found in
Anterior swim bladder extensions are well known from several members of teleost orders such as the clupeiform family Clupeidae (e.g.
So far, the function of the tissue pad is unknown. The pad might enhance the transmission efficiency from the swim bladder to the inner ears. The pad provides a larger contact region to the membranous labyrinth than the air-filled swim bladder extension alone which means that a larger area may transmit the pressure oscillations in the swim bladder wall to the perilymph surrounding the inner ear.
Our data clearly demonstrate that auditory sensitivities at frequencies above 0.3 kHz differ between the investigated cichlid species. The low absolute sensitivity and low maximum detectable frequency (0.7 kHz) in
Several morphological features of the swim bladder may influence the hearing abilities in fishes, particularly size, and distance to the inner ears.
In our study
The proximity of the swim bladder to the inner ears is likely to play an important role in transmitting sound efficiently to the inner ears. We showed that the closer the swim bladder came to the inner ears the higher the auditory sensitivities at frequencies above 0.3 kHz with exception of two out of eight frequencies in
There exists a discrepancy between holocentrids, sciaenids, and cichlids in the effect of the swim bladder-inner ear relationship on hearing sensitivity. This may partly be due to the fact that the swim bladder morphology and auditory sensitivities have not been investigated in the same study. Moreover, differences in the set-ups for AEP measurements may also make any comparison difficult.
The Holocentridae include species with unspecialized swim bladders (genus
In contrast to holocentrids, sciaenids with different swim bladder morphology differed mainly in their hearing bandwidth
It should be mentioned that fishes may possess enhanced auditory abilities without obvious morphological specializations. Damselfishes are sound-pressure sensitive and have best hearing sensitivities at 0.5 kHz without any swim bladder-inner ear connection
Different selective pressures may act on swim bladder size and the development of anterior extensions
It is possible that the swim bladder morphology depends on its function as a sound production organ. The cichlid species presently studied do not possess swim bladder drumming muscles and have not yet been studied with regard to sound production. A study on the sound production mechanism in the cichlid
Another set of important selective forces may have been ecological and ecoacoustical constraints. Rheophilic species such as
We conclude that anterior swim bladder extensions in cichlids result in higher auditory sensitivity at frequencies above 0.3 kHz. Large swim bladders appear to increase the maximum frequency detectable while anterior extensions increase the auditory sensitivity in a midfrequency region at 0.5 and 1 kHz.
We thank Dirk Schneider for providing