Abstract: The ability of cephalopods to hear has been controversial for some time (Moynihan 1985; Budelmann and Hanlon 1987). However, the organs specialized to detect tiny water vibrations characteristic of sound have been identified (Budelmann 1992), and experiments have been successful in showing that cuttlefish, Sepia officinalis, have the ability to associate low frequency sounds (below 100 Hz) with negative stimuli such as electric shock (Packard et al. 1990). In the following experiments, it was determined that cuttlefish can hear within a frequency range of 9 Hz to 9260 Hz, and that cuttlefish possess the ability to complete a complex task, combining the cognitive ability to associate a sound stimuli with the presentation of food. Introduction: Sound is generally defined as the transmission of mechanical vibrations through an elastic medium. In an aqueous environment, such acoustic energy results in rhythmic waves of compression and rarefaction of water molecules, as well as, particle displacement near the source (Kinne 1975). For cephalopods, perception of sound in the form of hearing has been a controversial subject (Moynihan 1985; Budelmann and Hanlon 1987). in part due to the difficulty in distinguishing between behavioral responses resulting from tactile vibration and sound perception. For mammals, the distinction is made simple by defining hearing as "the sense modality concerned with the perception of sound entering the ears at low and moderate intensities, as distinct from the perception of touch, vibration, and pain (Budelmann 1992). Such a definition would preclude cephalopods from having the ability to hear since they lack the developed "ear-like" structure being cited. However, Budelmann and Bleckmann (1988, see also Figure 1) describe the presence of epidermal lines similar to the lateral lines of fish and aquatic amphibians. In Sepia officinalis there are eight lateral lines" which extend from the dorsal surface of the head along the arms and contain mechanoreceptive hair cells that detect local water displacements. Budelmann also displayed that even tiny water displacements (2 um peak-to-peak displacement) along the epidermal head lines generated microphonic potentials indicating that these hair cells are indeed involved in the detection of water vibrations. However, another possible receptor involved in sound detection is the statocyst receptor system. Unlike the hair cells, these receptors are internal and greatly resemble the vertebrate inner ear. Statocysts are equilibrium receptors and can act as linear accelerometers detecting particle motion, since the entire animal vibrates with the water column during the production of sound (Budelmann 1992). A clear possibility is that both systems are involved. Several experiments have been conducted in which cuttlefish are shown to have the ability to associate sound with a stimuli. In one instance, Sepia was conditioned to the stimulus of sound vibrations (1 to 10 Hz) followed by mild electric shock. It was shown that an association could be made, and in later experimental sessions, fluctuations in breathing rhythms were observed to occur upon onset of the sound, before a shock was delivered (Packard et al. 1990). Although it thus seems that cuttlefish can detect sound stimuli, comparatively little work has been done testing frequencies above 200 Hz or attempting to condition cuttlefish to make à positive association between sound and a reward, such as food. In addition. adaptation or habituation to a particular sound occurs fairly rapidly and results in the loss of behavioral cues that indicate to the viewer that the sound is being heard. By conditioning the cuttlefish to associate the sound with a reward, the occurrence of a response becomes more reliable and frequencies and amplitudes may be altered to determine threshold hearing ranges. This paper addresses these ideas through three basic questions: (1) Within what range of frequencies and amplitudes can cuttlefish, Sepia officinalis, hear? (2) Can Sepia be conditioned to sound? (3) Can Sepia successfully complete a complex behavioral task that requires a higher level of cognitive association between sound and food? Materials and Methods: Animals: The experimental subjects included six cuttlefish, Sepia officinalis, (dorsal mantle length ranging from 2-7 cm) provided by the National Center of Cephalopods, Galveston, Texas and a large male cuttlefish (mantle length 23 cm) provided by the Monterey Bay Aquarium. Scanning Electron Microscopy: For SEM, one animal (20 mm mantle length) was pinned with a fine, insulated insect needle in a petri dish and fixed in 2% OsOA (Zetterquist buffer, pH 7.4) for 1 hour at 4°C. After fixation, the tissue was dehydrated in graded ethanol solutions, critical point dried in a Balzers CPD 020 unit, gold/palladium coated with a Hummer Il sputter unit (Technics, Alexandria, Virginia, USA), and viewed with a Hitachi S240 scanning electron microscope (Pacific Grove, California, USA). Behavioral Response Testing: In order to determine the frequencies at which cuttlefish are most responsive, a Lubell Model 98 hydrophone transmitter was lowered into a large circular tank (180 cm diameter, 57 cm deep, 12'C) along with a hydrophone receiver. As depicted in Figure 2A, bursts of sinusoidal signals produced by a B+K Precision 3020 Sweep/ Function Generator and amplified by a linear power amplifier were used to drive the submerged hydrophone. Both the stimulus sine wave and the output from the underwater hydrophone receiver were inputted into a Hitachi VC-6025 digital storage oscilloscope for sampling and comparison. The hydrophone output was also directly recorded onto videotape with a Canon Al Digital 8 mm video camera used for filming behavioral responses (shutter speed of 1/60th of a second). Frequencies ranging from 9-20,000 Hz were tested and behavioral responses of the large cuttlefish were recorded. The hydrophone produced short bursts of sound 1.15 seconds in length which occurred approximately every 2 seconds. Since adaptation to particular tones was found to rapidly occur, frequencies were varied quite often. An alteration in behavior upon onset of the tone was considered a behavioral response. Since the hydrophone was suspended by a rope, reaction to the sound was assumed to result from fluid motion rather than tank vibration. Conditioning Experiments: Two classical conditioning paradigms requiring different cognitive capabilities were designed in order to provide an independent test of hearing capabilities and to allow determination of threshold levels that would reliably prompt a specific behavior. The first conditioning experiment tested for an association between fifteen 180 Hz pulses of sound followed by the presentation of a crab. Three cuttlefish were placed in a plastic tank (4Ox 54 x 17 cm) with two 120 V fluorescent lights running along the sides of the tank. A four inch loud speaker was mounted on the external surface of the back wall of the tank; bolted and secured in place with silicon. This attachment allowed transmission of 180 Hz vibrations through the tank sides into the seawater. To control for tactile stimulation resulting from vibration of the tank, trials in which the cuttlefish were touching the sides or bottom of the tank were disregarded in the analysis. Each experimental session included three trials in which fifteen 180 Hz pulses 1.15 seconds in length were followed, approximately 5 seconds later, by the presentation of a crab through a long opaque tube. The entire set-up was enclosed by a black plastic drape to exclude visual stimuli. Each trial was recorded on video, through remote camera operation, and the tapes were later analyzed for behavioral responses. The second type of conditioning experiment had the goal of determining whether Sepia could associate a complex conditioned response with the presentation of sound. Whereas in the previous experiment behavioral responses were used as indicators that the sound was heard; this experiment required the cuttlefish to learn a task, associate it with a sound, and then carry out the task upon initiation of the sound. Two animals were used for this test. They were placed in a glass tank (SOx19x15 cm) and maintained at an approximate temperature of 14'C. A clear Plexiglas barrier was positioned at a 45' angle to the walls in one corner partially sectioning off that corner of the tank but allowing a passageway between the barrier and one wall (see Figure 2B). Black plastic netting (0.8 cm mesh size) was glued on the surface of the Plexiglas to provide texture and a visual cue as to where the barrier ended. The opaque tube used for the introduction of crabs into the tank was positioned in the comer behind the barrier. The hydrophone used in this case was made from a four inch stereo speaker secured tightly within a Tupperware container, and held on the surface of the water. Initial sessions consisted of fifteen 180 Hz pulses (1.15 seconds in length) being followed after 5 seconds with the presentation of a crab behind the barrier. During this period, the crab could be seen through the netted barrier, so the cuttlefish could associate the sound with the visual stimulus of food and the task of maneuvering around the barrier. After 25 sessions it was clear that Sepia had mastered this task, and the barrier was covered with black, opaque plastic. Any resulting movement by the Sepia around the barrier upon initiation of the sound could, therefore, only be a result of the sound stimulus and the learning associated with it. Results: Scanning Electron Microscopy: By using scanning electron microscopy to examine the dorsal head and arm region of the cuttlefish, the epidermal head lines characterized by Budelmann and Bleckmann (1988) were viewed. Each line consisted of a groove containing hair cells that possessed many kinocilia (up to 12 um in length). The hair cells did not form a continuous line; single hair cells and groups of hair cells occurred in alternation (Sundermann 19883). Behavioral Responses to Sound Presentation: Although a range of frequencies from 9 Hz to 20,000 Hz was tested (limited by the capabilities of the hydrophone), responses were only observed for frequencies between 9 and 9260 Hz. All responses were divided into two categories; startle and non-startle behavioral responses. Startle responses include color flashes, head retractions, or increasing fin beats and occur immediately (within 3-8 video frames) at the onset of the first sound pulse. Non- startle behavioral responses are distinguishable from startle responses in that the reactions are more prolonged, appear to be controlled, and are not subject to habituation. Such responses include rotation or motion toward the source of sound production, fin beating in synchrony with sound pulses, avoidance responses, and color changes. High amplitude tones over a large frequency range were successful in eliciting startle responses (see Figure 3), and a "threshold" level exists at an intensity of approximately 5.5 V, above which all responses occur in connection with a startle response. The most common reaction to such high amplitude sounds was a slight drawing in of the tentacles, flaring of the fin, and retraction of the head. Low amplitude signals produced startle responses only occasionally, but a variety of continuous behavioral responses were produced, again with sounds of a wide frequency range. An unusual reaction was produced to a stimulation of 9 Hz. On six separate occasions, the cuttlefish turned or proceeded in the direction of the hydrophone upon initiation of the 9 Hz tone. Conditioning Experiments: The first conditioning experiment attempted to determine whether Sepia would display excitatory behavior during sound production in anticipation of food. The resulting responses are summarized in Figure 4 and correspond to the reactions that preceeded the presentation of the crab. Color changes were the first continuous behavioral responses to be observed and were normally seen as a darkening of the chromatophores throughout the body or tentacles, but the incidence of color change decreased with the number of trials. Fin beating, defined as an increase in beating rate or beating in synchrony with the sound pulses, also was regularly displayed after approximately 10 trials. Rotation or movement toward the tube before the introduction of a crab increased with the number of trials. The final factor used to deduce anticipatory behavior was the positioning in an attention posture (Messenger 1973), with two tentacles raised. This behavior is normally seen during preparation for prey attack. It was virtually absent during the first fifteen trials, after which its occurrence became more regular. Figure 5 shows the percentage of trials, out of the total number of trials, in which each benavioral response occurred. Changes in the rate of fin beating was seen most frequently, in 51 percent of the observations. The second conditioning experiment was more complex and required the cuttlefish to not only associate the sound with a reward (crab), but also to complete a task in order to obtain the reward. The task was considered to be successfully completed only if the cuttlefish displayed preparatory behavior by moving behind the barrier before the crab was introduced. To discount the possibility of reaction towards other stimuli (including vibration caused by the entrance of the crab into the water, chemical stimuli, etc.), trials in which such movement occurred after the crab was dropped into the tube were not considered successes. Figure 6 shows that the cuttlefish became fairly competent in obtaining the crabs early in the training period. This was accomplished by either moving around the barrier after the crab was presented or waiting until the crab had moved into view in the main tank compartment. Successful task completion did not occur until the fourteenth trial, but occurred rather frequently thereafter (7 out of 19 trials) Discussion: Despite testing over a wide range of frequencies between 9 and 20,000 Hz at amplitudes up to 8.5 V, no responses of any kind were seen above 10,000 Hz, and responses above 5,500 Hz were rare. This indicates that, while the hearing range is broader than previously known (Packard et al. 1990), the highest occurrence of responses occurred at relatively low frequencies. In humans, sensitivity is much greater for frequencies that correspond to the human voice (approximately 1000 Hz). Similarly, in cuttlefish, there may be an adaptive value of hearing low frequencies characteristic of prey or predator tail beats or other sounds associated with movement. In addition, sensitivity to low frequencies could be useful for communication. Plainfin midshipman fish, Porichthys notatus, produce low frequency sounds by contraction of the intrinsic muscles of the swimbladder (Cohen and Winn 1967). A specific frequency range of 98-108 Hz has been shown to elicit male-searching behavior in gravid females (Ibara et al. 1988), indicating that production and reception of such sounds is essential in communication. Figure 3 displays a fairly distinct behavioral response line at 5.5 V, above which all non-startle behavioral responses occurred in association with a startle response. Thus, an amplitude of 5.5 V could possibly represent the threshold for physiologically relevant sound intensities in nature, suggesting that intensities above this point are not normally experienced. To determine whether such an assumption is true, studies would have to be undertaken which examine the sound intensities and frequencies characteristically produced by oceanic species. Sepia also displayed the ability to locate the source of sound production. In the set-up, two foreign objects, the hydrophone and the hydrophone receiver, were introduced into the tank. On nine separate occasions, the cuttlefish displayed movement or rotation towards the hydrophone upon initiation of sound. However, there were no similar reactions toward the hydrophone receiver suggesting that cuttlefish can detect the directionality of sound production and can quite accurately locate its source. In addition, the fact that Sepia displays a particular sensitivity to low frequencies and can associate sound with food raises the question of whether such an ability is useful during prey capture. For this to be true, the cuttlefish would have to associate a particular frequency with the prey. The results of this experiment clearly display that such an association is possible. Referring once again to Figure 4, the increasing incidence of preparatory activities such as rotation or movement toward the tube or assumption of the attention position, with tentacles raised, indicates successful learning; associating sound with prey. A somewhat unusual result was the decline in frequency of color changes. Since visual color displays are used for intraspecific communication in Sepia (Tinbergen 1939), a possible explanation for this decrease in color changes arises from the fact that the three animals resided within one tank. There was, therefore, competition for the food since a single crab was presented during each trial. Color changes not only indicate to the viewer that the sound was heard, but could also conceivably arouse the interest of the other animals. Therefore, it is in the best interest of the cuttlefish to control such blatant responses, increasing their own chance of obtaining the prey. A single animal in a tank would be a useful control in answering such a question. The goal of the final experiment was to determine whether cuttlefish could master a task demanding higher learning capabilities which required sound discrimination. Proficiency in completing the task required a level of cognition beyond the simple association between sound and food displayed in the previous experiment. For successful execution of the task, the cuttlefish had to make the connection between sound, food, and movement around the barrier. As seen in Figure 6, motivation, defined as the desire to obtain food either by procession around the barrier before or after crab presentation or by waiting for the crab to escape, remained fairly constant throughout the experiment. This is important in showing that, although the interest was present at the start of the experiment. fourteen trials were required before the cuttlefish learned to successfully complete the task. The relatively high number of successful responses after trial fourteen signifies a clear ability to leam a complex task while making the association between sound and food. Thus, the questions posed at the start of the investigation were answered. Cuttlefish can hear within a wide range of frequencies, displaying particular sensitivity to low amplitude, low frequency sounds. In addition, cuttlefish possess the ability to leam a complex task through association of sound with a reward. Acknowledgments: T am grateful to Dr. William Gilly, Professor of Neurobiology at Hopkins Marine Station, for being supportive throughout the project, for calming my frustrations, and for being an inspiration as a creative research scientist. My thanks, also, to Dr. Thomas Preuss for his enduring enthusiasm for the project, constant support, and numerous trips to the hardware store. In addition, I would like to thank Dr. Jim Watanabe for use of his hydrophone and for his undying dedication to our class as a whole, Dr. Mark Denny for his help in coming up with creative solutions for an underwater sound device, John Lee for his electrical expertise in building the sound devise, and the Weissmann lab for our use of their tanks. References: Budelmann, B.-U. and Hanlon, Robert T. February 1987. Why Cephalopods Are Probably Not "Deaf". The American Naturalist. Vol. 129; No. 2; p. 312-317. Budelmann and Bleckmann. 1988. A lateral line analogue in cephalopods: water waves generate microphonic potentials in the epidermal head lines of Sepia and Lolliguncula. Journal of Comparative Physiology A. Vol. 164; p. 1-5. Budelmann, Bernd U. 1992. Hearing in Nonarthropod Invertebrates. The Evolutionary Biology of Hearing. Springer-Verlag, New York. Ch. 10; p. 141-155. Cohen, M.J. and Winn, H.E. 1967. Electrophysiological observations on hearing and sound production in the fish, Porichthys notatus. Journal of Experimental Zoology. Vol. 165; p. 355-370. Ibara, Richard M., Larry T. Penny, Alfred W.Ebeling, Gilbert van Dykhuizen, and Gregor Cailliet. 1983. The mating call of the plainfin midshipman fish, Porichthys notatus. Predators and prey in fishes. p. 205-212. Kinne, O. 1975. Marine Ecology: A Comprehensive, Integrated Treatise on Life in Oceans and Coastal Waters. John Wiley and Sons; Vol. II, Part 2; p. 740-742. Messenger, J.B. 1973. Learning in the cuttlefish, Sepia. Animal Behavior. Vol. 21; p. 801-826. Moynihan, Martin. March 1985. Why are Cephalopods Deaf? The American Naturalist. Vol. 125; No. 3; p. 312-317. Packard, A.; Karlsen, H.E.; Sand, O. 1990. Low frequency hearing in cephalopods. Journal of Comparative Physiology A. Vol. 166; p. 501-505. Sundermann, G. 1988. The fine structure of epidermal lines on arms and head of postembryonic Sepia officinalis and Loligo vulgaris (Mollusca, Cephalopoda). Cell Tissue Res. Vol. 232; p. 669-677. e Tinbergen, L. 1939. Zur Fortpflanzungsethologie von Sepia officinalis. L. Archives Neerlandaises de Zoologie. Vol. 3; p. 323-364. Figure 1: A. Scanning electron micrograph of the dorsal head and arms of the cuttlefish, Sepia officinalis. The lateral lines containing hair cells can be seen running from the head, down the arms. B. A single ciliated hair cell within the lateral line channel. Figure 2: A. Set-up used in determining the frequency and amplitude ranges for cuttlefish, Sepia officinalis. B. Set-up used in the second conditioning experiment, displaying the barrier that the cuttlefish had to proceed around in order to obtain the food. Figure 3: Responses to frequencies between 9 and 10,000 Hz and amplitudes of 8V to 8.2V. Startle responses refer to those reactions that occur during the first burst of sound. Behavioral responses involve more sophisticated reactions to sound including; rotation or movement toward the source, fin beating in synchrony with sound pulses, escape responses, or long term color changes. Figure 4: Distribution of responses during conditioning to 180 Hz. Color changes, increased or synchronous fin beating, rotation or movement toward the tube and assumption of the attention position, with two tentacles raised, were all considered behavioral responses to the sound pulses. Rotation or movement toward the tube and raising of the tentacles are the clearest indicators that an association between sound and the presentation of food has been made. Figure 5: The percentage of trials in which particular behavioral responses to 180 Hz sound pulses were observed. Figure 6: c T'ime course for learning experiment. Successful task completion involves only those trials in which Sepia traveled around the barrier before the crab was presented. B. Figure 1: A. Li 1 NOR Figure 2: A. Receiver 10 Hydrophone Transmitter S Hydrophone Barrier — Oscilloscope A Function Generator Power Amplifier 50 (o uI 8 o 8. 23 8 E. 8 fegt 9 8 8 (%) Suosgo S 1 Oo OH H H -6 8 8