Sound Localization in Catfish
Fishes have developed a wide diversity of inner ears and peripheral structures facilitating sound transmission to the inner ears. The significance of this morphological diversity in ear structure however, is widely unknown. Several fish taxa possess functional peripheral structures, which conduct sound vibration to the inner ears and serve to enhance their hearing sensitivity. Otophysi for example, developed "middle ear'-like ossicles (Weberian ossicles) (Frisch, 1938) to transmit vibrations of the swim bladder to the inner ear, working as a tympanic membrane. It has been shown that Weberian ossicles enhance hearing in goldfish 30 dB more, depending on frequency (Ladich and Wysocki 2003).
Fishes in the superorder Otophysi are mostly freshwater teleosts that include catfishes, knifefishes, minnows, carps, characins, suckers and loaches (Rosen & Greenwood, 1970). In the catfish family Loricariidae, the Weberian apparatus is minimized, since the swim bladder is directly adjacent to each side of their ears (Aquino and Schaefer, 2002). Loricariid catfishes feature a highly derived swim bladder morphology characterized by complete division of the bladder into two separate spheres that are each surrounded by a megaphone-like bony capsule. These catfishes also have an unusual pterotic bone, which is adjacent to their bi-lobed swim bladder that has holes in the skull region called fenestrae which also aid in sound localization. (Weitzmann, 2005).
Purpose of the project:
Catfishes reveal a high diversity in the morphology of the swim bladder and Weberian ossicles. Some possess large unpaired bladders whereas others have small, paired, encapsulated bladders. Catfishes with free bladders hear significantly better above 1 kHz than those having encapsulated bladders (Lechner and Ladich, 2008). However, little is understood about either the morphological range of these structures in the family Loricariidae, or about their acoustical functionality.
We hypothesized that having a bi-lobed swim bladder provides Loricariid catfishes with an improved ability to localize sounds. In order to test this hypothesis, we first had to develop a protocol for attracting Otocinclus affinis, a small loricariid catfish that is found in schools of several thousand about 1cm to 3 cm in length found primarily in the Amazon River basin, to a sound source. We did this by using operant conditioning, in which a food reward was paired with a conspecific sound stimulus (Fay, 1988).
Like many catfishes (Ladich and Fine,2006), we have also found that O. affinis produces sounds by stridulating their pectoral spines to make broad-band clicks (unpublished data). Although these sounds are produced when the fish are handled, the normal behavioral context of these sounds are currently unknown. Hypothesized behavioral contexts are various agonistic interactions, courtship, spawning and competitive feeding.
Following operant conditioning of O. affinis to a conspecific click sound, we then set out to see if O. affinis were attracted to a non-conspecific tone. We used a 500-hz tone to test this, following all the same parameters of the operant conditioning. Once we achieved above 70% success rate, we then used a multiple speaker array to examine the ability of O. affinis to localize to a sound source in a large, round tank by setting up 4 equal quadrants with 1 speaker along the wall in the middle of each quadrant and the release container in the middle intersecting points of the 4 quadrants. By allowing the O. affinis to acclimate in the release container for 5 minute and randomly choosing a speaker to play the sound through we were able to visually see the O. affinis could localize sound. To further test whether the bi-lobed swim bladder had an effect on the O. affinis we recorded the results of the trained O. affinis then we deflated the swim bladders and subject the deflated O. affinis through the same 4 speaker test, and thus were able to see that the deflated O. affinis had difficulty in localizing sound.
The three purposes of the project are summarized below:
- Train Otocinclus affinis using operant conditioning to approach a conspecific sound source.
- Test the acuity of sound localization in trained O. affinis.
- Examine the importance of the swim bladder in sound localization.
Subjects & Training
Otocinclus affinis, ranging in total length between 1.5 and 2 cm, were kept in a 122 x 30 cm tank with water temperature at approximately 26.5 °C. Two types of experiments were performed- training and test. During training trials (N=3), food (algae pellets) and a sound stimulus were presented simultaneously. Two groups of 42-50 fish each were trained to associate the sound stimulus with food by giving them food and sound stimuli simultaneously three times a week for 6 weeks. The naïve group (N=42) only received food during this period. Figure 1 presents the groups of fish, trained and naïve used in the experiments of this study.
Experiments & Video Analysis
One-speaker experiments: For each treatment, groups of fish were moved into an experimental aquarium and were allowed to acclimate for 1 week. Treatments were presented as shown above (Fig. 1). The sound stimulus used in this experiment was a 4-sec recording, played continuously, with a single conspecific O. affinis click. Sound was played through a computer to an Audiosource power amplifier to a University Sound UW-30 underwater speaker (Fig. 2A). For each trial, fish behavior was video-recorded for 15 min before, during, and after the sound stimulus. Sample frames of video images were analyzed every 3 min of each 15 min recording period. At each frame, numbers of fish were counted in each column of a 15x15 cm grid drawn on the back of the aquarium. The locations of the fish in the aquarium were monitored throughout the course of the experiment using a Sony HandyCam with a wide-angle lens in night vision mode. The aquarium was illuminated by an infrared light to prevent disturbance of the fish, which are nocturnal.
Four-speaker experiments: For each trial (N=26), an individual "conditioned" fish was moved into a 1.3 m diameter round tank (Fig. 2B) and was allowed to acclimate in a clear plastic cylinder for 5 min. The cylinder was then lifted, allowing the fish swim freely, and the conspecific sound stimulus was then presented for 5 minutes. The final 5 min was a post-stimulus period, in which no sound was administered. The sound stimulus used was the same as that for the one-speaker experiments (Fig. 3B), with sound being emitted by one randomly chosen UW-30 speaker of four mounted at 0, 90,180, and 270° along the edge of the tank. The locations of the fish in the aquarium were monitored throughout the course of the experiment using a Sony HandyCam with a wide-angle lens. Sample frames of video images were analyzed for each 5 min recording period, in which numbers of fish were counted in each quadrant of the aquarium.
Four-speaker experiments with deflated swim bladders: For each trial (N=9), an individual "conditioned" fish was moved into a 1.3 m diameter round tank (Fig. 2B) and was allowed to acclimate in a clear plastic cylinder for 5 min. The cylinder was then lifted, allowing the fish swim freely, and the conspecific sound stimulus was then presented for 5 min. The final 5 min was a post-stimulus period, in which no sound was administered. Then the "conditioned fish" were taken for swim bladder deflation, and returned to acclimate in the test tank. Within 2-3 h, the deflated, conditioned fish (N=5), were then allowed to acclimate in a clear plastic cylinder for 5 min. The cylinder was then lifted, allowing the fish swim freely, and the conspecific sound stimulus was then presented for 5 min. The final 5 min was a post-stimulus period, in which no sound was administered. There were four fish that did not survive the deflation process. The sound stimulus used was the same as that for the one-speaker experiments (Fig. 3B), with sound being emitted by one randomly chosen UW-30 speaker of four mounted at 0, 90,180, and 270 along the edge of the tank. The locations of the fish in the aquarium were monitored throughout the course of the experiment using a Sony HandyCam with a wide-angle lens. Sample frames of video images were analyzed for each 5 min recording period, in which numbers of fish were counted in each quadrant of the aquarium.
In figure 4, we present graphically the results of our operant conditioning test, where following training with a paired food/sound stimulus, Otocinclus affinis were significantly attracted (p<0.05 when tested against nave controls) to the right side of the tank with sound alone. Although conditioning was done with conspecific clicks, fish were attracted to both conspecific sounds (Fig. 4B) and a 500 Hz tone (Fig. 4C). To check for the internal validity of our operant conditioning test, we observe that the conspecific sound by itself did not significantly attract O. affinis, suggesting that the fish were responding only to the conditioning to a food stimulus.
Testing for sound localization, (figure 5) we see that the total time that otocinclus spent in each tank quadrant was dependant to a function of which speaker was emitting the conspecific sound stimulus. At the beginning of each sound localization trial, 19 out of 26 fish (73%) swam directly to the quadrant of the correct sound-producing speaker first. Fish spent significantly more time near the sound-producing speaker than the other three speakers (p <0.05). This was true for all four speakers. Also, fishes were attracted to the area of the speaker even after the stimulus stopped (data not shown).
Following, swim bladder deflation was performed to test if the loss of their functioning ability would affect the fishes' ability to localize sound. The X-rays in figure 2C show the deflated swim bladders that appear opaque on the radiology film. In figure 6, we observe the time spent at the correct stimulus speaker and the correlation between control and deflated fish. Of the 24 fish included in this experiment, 14 were used for control out of whom, 100% went to the correct speaker initially. Out of the 10 deflated fish included, 0% went to the correct speaker initially. O. affinis therefore, lost their ability to localize sound once the swim bladders were deflated, suggesting that this taxa of fish use their swim bladder to help localize sound properly.
It is anticipated that future experiments will use this conditioning paradigm, to further study the sound localization anatomy in O. affinis. In figure 7, we see the percent of O. affinis that swam first to correct quadrants (with sound producing speakers) or incorrect quadrants, as a function of swim bladder deflation and the percent time spent in the correct quadrant as a function of swim bladder deflation. Fishes with deflated swim bladders never initially localized to the correct quadrant and rarely spent time in that quadrant.
Fishes have an astonishing diversity in sound generating and acoustic functioning structures; this diversity is not only found between different orders but is also expressed in varying magnitude between members of the same order. Several studies have tried to experimentally and genetically elucidate the ontogeny (Coburn, 1998) and in particular the role of CNS development in the diversity of the acoustic machinery in fish (Lechner, Wysocki and Ladich ,2010)( Wysocki and Ladich ,2001)( Ladich and Popper, 2004).
Swim bladders have long been found to contribute as sound producing organs in fishes. Sound production is achieved by rapidly contracting muscles to vibrate the swim bladder and pectoral girdle and by bony elements, which produce sound by stridulation or plucking tendons (Ladich and Fine 2006). Swim bladder muscles of fishes can contract at frequencies up to 250 Hz, making them the fastest muscles in the animal kingdom. The small tensor tripodis muscle (TT) contractions most likely prevent transmission of swimbladder vibrations to the inner ear via the Weberian ossicles during vocalization. (Ladich, 2001). The recognition however of the role of these muscles and structures in sound localization has not been demonstrated.
Acoustical communication is sought to arise from the collaboration of sound producing and acoustic structures, however, the degree of correlation between sound production and hearing in fishes remains unknown. In order to determine whether fishes are able to utilize temporal characteristics of acoustic signals, previous experiments have determined time resolution in otophysines and anabantoids by analyzing auditory evoked potentials (AEPs) to double-click stimuli with varying click periods (Wysocki and Ladich, 2002). Moreover, temporal recognition of sound was found to have wide diversity among members of the Loricariidae (Aquino and Schaefer 2002).
A localization mechanism that exploits the amplitude, time, or phase difference between the ears as employed by earthly vertebrates is not available to fish because the ears are very close together, the speed of sound in water is more than three times faster than in air, and the close impedance match between the fish's body and water precludes usable diffracted paths. It is suggested that a quadrupole mechanism (the multipole hearing theory) exists that is used by hair cells lacking an overlying otolith as in the Loricariidae (Roger and Zeddies, 2008).In the grounds of this hypothesized mechanism lies the orientation theory, where the sensory epithelia (maculae) of otolithic end organs contain groups of hair cells with the same or similar morphological orientation, which are typically oppositely oriented across a dividing line. The utricle and lagena of most fishes follow a conserved pattern but the saccule is more varied (Popper and Fay 1999; Popper et al. 2003; Popper and Schilt, Chapter 2). This mechanism can explain the directional hearing of fishes, but it cannot further shed light on the sound localization and orientation as the one observed in our experiments.
The ability of the auditory system to process fish's sounds (Wysocki and Ladich 2005) is also investigated in connection to the effects of noise in the aquatic environment. These studies have shown that noise exposure (white noise of 158 dB re 1 lPa for 24 h) negatively affects the detection of short transient signals and the temporal resolution ability in fishes. It is true that high-intensity sound can change the behaviour of animals, cause severe damage to inner organs and induce endocrinological stress responses. Enger (1981) observed damage to the sensory epithelia of the inner ear of the cod Gadus morhua after exposure to intense noise. However, fishes are able to recover from noise-induced threshold shifts within a few days, probably due to hair cell recovery or regeneration capabilities (Smith, 2006). In our study, the maximum continuous sound exposure was 5 minutes, safely allowing us to conclude that no sensory damage was accidentally inflicted to the fishes, in the course of any of the experiments performed. Moreover, sounds perceived by fish as biological are usually at a lower frequency level (Rogers and Cox 1988).
Spatial discrimination has been long suggested to occur in fishes by the use of receptors arrayed over a large area of the fish skin, which act like a retina upon which the discharge projects "electric images" (the electrolocation process) (Graff et al, 2004). This mechanism is sought to facilitate distant discrimination of objects according to their composition, their size or their distance and is sought to arise in catfish that are not, however, conditioned to the administration of sound and feeding. In our experiments, we have overcome the disambiguation of this mechanism in our hypothesized mechanical/sensory sound localization by swim bladders, by randomly using the chosen UW-30 speaker out of four mounted at 0, 90,180, and 270 along the edge of the tank. However, it can be argued furthermore that with reward and training, the fish could possibly recognize the sound origination by recognition and conditioning of piezoelectric stimuli or even changes in surface water wave formations (infrasound and linear acceleration-Sand and Karlsen, 2000) sent off by the speakers. It is true that underwater speakers inside small tanks generate high amounts of particle motion relative to pressure (Parvulescu, 1967). This limitation is overcome when placing the fish 3-50 cm below the water surface, as we did in our experiments, when using an underwater source (Smith, 2006),(Mann et al, 2001). This is clear because the lateral line system is able to detect nearfield particle motions only up to at a distance of a few centimeters (Sand 1981).
Moreover, even if the pressure/particle motion hypothesis was true, the finding of loss of localization when swim bladders were deflated would further prove that swim bladders are the exact and unique location of the sound localization mechanism in Loricariidae. The electrophysiological data on directional sensitivity obtained by vibrating the goldfish in air by Ma and Fay, 2002 exclude pressure stimulation via the swim bladder. However, a separate encoding of sound pressure and incident particle motion, and phase comparison of these sound parameters is possible in the swim bladder. The otophysan species ide (Leuciscus idus) is an example, being able to discriminate between opposing sound sources in the horizontal plane (Schuijf et al. 1977). In Atlantic cod, an auditory function of the swim bladder has been demonstrated in both behavioral (Chapman and Hawkins 1973) and electrophysiological (Sand and Enger 1973) experiments.
ABR or auditory brainstem response is a non-invasive far-field recording of synchronous neural activity in the auditory pathway and has been shown to be superior to training since it includes rapid whole-animal measurements without time-consuming training and the opportunity to repeatedly use the same test subjects since it does not permanently harm animals. We have chosen not to use ABR in the current experiments but we anticipate their application in future studies directed at the electrochemical proof of spatial localization in O. affinis using swim bladder impaired fish subjects, to further elucidate the neuronal mechanisms behind the control of spatial localization.
In conclusion, the spatial localization of O. affinis was shown experimentally to depend upon the intact condition and functional ability of the swim bladder, which when deflated demonstrated the fishes' reduced capability to locate a previously identified sound source (conditioned with feeding). This finding is novel in fish bioacoustics and represents initial experimental proof in the characterization of this anatomical structure (the Weberian ossicle) as a middle-ear pre-evolutionary structure (Chardon, 1997) , which provides with sense of direction of sound and serves as a potential bone or sensory transducer for the upper level neuronal discrimination in brainstem auditory and perception centres. Further research is warranted for the experimental replication of the findings, and the elucidation of all the involved steps in spatial localization in O. affinis.
Aquino A. E.; Schaefer S. A. 2002. The temporal region of the cranium of loricarioid catfishes (Teleostei: Siluriformes): morphological diversity and phylogenetic significance. Zool. Anz. 241: pp. 223-244, 2002.
Arratia G. 2003. Catfishes, Vol 1. Enfield, NH: Science Publishers, Inc.
Bridge TW, Haddon AC. 1889. Contribution to the anatomy of fishes volume I. The airbladder and Weberian ossicles in the Siluridae. Proc Roy Soc Lond 46: 209-227.
Chapman CJ, Hawkins AD 1973 A field study of hearing in the cod, Gadus morhua L. J Comp Physiol 85:147-167
Chardon M, Vandewalle P. 1997. Evolutionary trends and possible origin of the Weberian apparatus. Neth J Zoo 47 (4): 383-403
Coburn MM, Grubach PG. 1998. Ontogeny of the Weberian apparatus in the armored catfish Corydoras paleatus (Siluriformes: Callichthyidae). Copeia 1998: 301-311
Dijgraaf S. 1960. Hearing in Bony Fishes. Proc Roy Soc Lond. Series B. Vol. 152. 946: 51-54
Enger, P.S., 1981. Frequency discrimination in teleosts - central or peripheral. In: Tavolga, W.N., Popper, A.N., Fay, R.R. (Eds.), Hearing and Sound Communication in Fishes. Springer, New York, pp. 243-253.
Fay R. R.1988 Hearing in Vertebrates: a Psychophysics Databook. Hill-Fay Associates, Winnetka, IL, USA, 1988.
Fay R. R. 2005 Sound source localization by fishes. In A. N. Popper and R. R. Fay, eds., Sound Source Localization, pp. 36-66. Springer Science - Business Media, Inc., New York, N.Y., 2005.
Godin OA. 2008. Sound transmission through water-air interfaces: new insights into an old problem. Contemp Phys. Vol. 49. (2) p105-123
Graff C., Kaminski G., Gresty M., Ohlmann T. 2004. Fish Perform Spatial Pattern Recognition and Abstraction by Exclusive Use of Active Electrolocation Current Biology, 14(9); p: 818-823
Ladich F, Popper AN. 2004. Parallel evolution in fish hearing organs. In: Manly GA, Popper AN, Fay RR, editors. Evolution of the vertebrate auditory system. New York, NY: Springer. p 95-127.
Ladich F, Fine M. 2006 Sound-generating mechanisms in fishes: a unique diversity in vertebrates. In F. Ladich, S. P. Collin, P. Moller, and B. G. Kapoor, eds., Communication in Fishes, Vol. I., ch. 1, pp. 3-43. Science-Publishers, Enfield, N.H., 2006.
Ladich F, Wysocki LE. 2003. How does tripus extirpation affect auditory sensitivity in goldfish? Hear Res 182:119-129.
Ladich, F. 2001. The sound-generating (and -detecting) motor system in catfishes: design of swimbladder muscles in doradids and pimelodids. Anatomical Record 263, 297 - 306
Lechner W, Ladich F. 2008. Size matters: diversity in swimbladders and Weberian ossicles affects hearing in catfishes. Jour Exp Biol 211: 1681-1689
Lechner, W., Wysocki L.E., Ladich F. 2010. Ontogenetic development of auditory sensitivity and sound production in the squeaker catfish Synodontis schoutedeni. BMC Biology 8, 10
Lombarte A, Popper AN. 2004. Quantitative changes in the otolithic organs of the inner ear during the settlement period in European hake Merluccius merluccius. Mar Ecol Prog Ser 267: 233-240.
Ma WLD, Fay RR 2002. Neural representations of the axis of acoustic particle motion in nucleus centralis of the torus semicircularis of the goldfish, Carassius auratus. JComp Physiol A 188:301-313.
Mann, D.A., Higgs, D.M., Tavolga, W.N., Souza, M.J., Popper, A.N. 2001. Ultrasound detection by clupeiform fishes. J. Acoust. Soc. Am. 109, 3048-3054.
Parvulescu, A., 1967. The acoustics of small tanks. In: Tavolga, W.N. (Ed.), Marine Bio- Acoustics. Pergamon Press, Oxford, pp. 7-13.
Popper AN, Fay RR, 1999. The Auditory Periphery in Fishes. In: Fay RR, Popper AN, editors. Comparative hearing: fish and amphibians. New York: Springer. p 43-100.
Popper AN, Schilt CR. 2008. Hearing and acoustic behavior: basic and applied considerations. In: Webb JF, Popper AN, Fay RR, editors. Fish Bioacoustics. New York, NY: Springer. p 17-48.
Rogers P, Cox M. 1988. Underwater sound as a biological stimulus. In: Atema J, Fay RR, Popper AN, Tavolga WN, editors. Sensory Biology of Aquatic Animals. New York, NY: Springer-Verlag, p 131-149.
Rogers PH., Zeddies DG. 2008. Multipole Mechanisms for Directional Hearing in Fish. In: Webb JF, Popper AN, Fay RR, editors. Fish Bioacoustics. New York, NY: Springer. p 99-144.
Rosen DE, Greenwood PH. (1970). Origin of the Weberian apparatus and the relationships of the ostariophysan and gonorynchiform fishes. Amer Mus Nov 2428: 1-25
Saidel WM, Popper AN. 1983. Spatial organization in the saccule and sagena of a teleost: hair cell pattern and innervation. J Morph 177: 301-317.
Sand O. Bleckmann H. 2008. Orientation to Auditory and Lateral Line Stimuli. In: Webb JF, Popper AN, Fay RR, editors. Fish Bioacoustics. New York, NY: Springer. p 99-144.
Sand O, Enger PS 1973. Evidence for an auditory function of the swim bladder in the cod. J Exp Biol 59:405-414.
Sand O, Karlsen HE.(2000) Detection of infrasound and linear acceleration in fishes. Philos Trans R Soc Lond B Biol Sci. 29;355(1401):1295-8.
Sand O 1981 The lateral-line and sound reception. In: Tavolga WN, Popper AN, Fay RR (eds) Hearing and Sound Communication in Fishes. New York: Springer-Verlag, pp. 459-480
Schaefer S. 1987. Osteology of Hypostomus plecostomus with a phylogenetic analysis of the loricariid subfamilies (Pisces: Siluroidei). Contrib Sci 394:1-31
Schuijf A, Visser C, Willers AFM, Buwalda RJA (1977) Acoustic localization in an ostariophysian fish. Experientia 33:1062-1063.
Smith, M.E., Coffin, A.B., Miller, D.L., Popper, A.N., 2006. Anatomical and functional recovery of the goldfish (Carassius auratus) ear following noise exposure. J. Exp. Biol. 209, 4193-4202.
Van Bergeijk WA. 1967. The Evolution of Vertebrate Hearing. In: Neff WD, editor. Contributions to Sensory Physiology. Vol. 2. New York, NY: Academic. p 1-49.
von Frisch K.(1938) The sense of hearing in fish. Nature 141: pp. 8-11, 1938.
Weber EH. 1820. De Aure et Auditu Hominis et Animalium. Pars I. De Aure Animalium Aquatilium. Leipzig: Gerhard Fleischer.
Weitzmann SH. 2005. Hearing in catfishes especially that of the family Loricarridae. In: Hans-Georg E, Seidel I, editors. Catfish Atlas. Vol. 1. Washington DC: Baensch. P 31-39.
Wysocki, L. E., Ladich F. 2001. The ontogenetic development of auditory sensitivity, vocalization and acoustic communication in the labyrinth fish Trichopsis vittata . Journal of Comparative Physiology A 187, 177 - 187