Association between burst-pulse sounds and aggressive behavior in captive Atlantic bottlenosed dolphins (Tursiops truncatus)

Zoo Biology 2:93-103 (1983)

Association Between Burst-Pulse Sounds and Aggressive Behavior in Captive Atlantic Bottlenosed Dolphins (Tursiops truncatus) Neal A. Overstrom

Sea Research Foundation, Inc, Mystic Marinelife Aquarium, Mystic, Connecticut

A specific display observed during agonistic behavior among captive Atlantic bottlenosed dolphins (Tursiops truncatus) was examined. The primary component of the display was an open-mouthed posture accompanied by violent vertical head motions and the emission of pulse-type vocalizations. Jaw-clap behavior produced during the display was quantified for use as an index of aggressive motivation. By aurally monitoring the animals’ vocalizations it was found that the level of aggressive response between the participants of the interactions increased with the production and subsequent duration of burst-pulse sounds. The possibility of burst- pulse sounds resulting in auditory or tactile discomfort when directed toward conspecifics is discussed.

Key words: Atlantic bottlenosed dolphin, (Tursiops truncatus), agonistic interaction,

INTRODUCTION

vocalization, burst-pulse emission, aggression, captivity, social behavior

In 1978 Myers and Overstrom described a specific interaction occurring among two male and two female Atlantic bottlenosed dolphins (Tursiops truncatus) held in captivity at Mystic Marinelife Aquarium in Mystic, Connecticut. The behavior oc- curred between training demonstrations while the animals were segregated in separate quadrants of a holding pool by a carousel-style divider. Two animals from adjacent quadrants, usually the males, would rush together until facing each other but separated by a net (Fig. 1). While in this position one or both would make vertical head motions, simultaneously expelling bubbles and emitting bursts of broad-band pulsed sounds (burst-pulse emissions) frequently referred to as “squawks” in the literature (see Popper [1980] for a review of delphinid sounds). Usually the participants’ mouths

Received for publication November 19, 1982; accepted December 27, 1982. Address reprint requests to Neal A. Overstrom, Education Department, Mystic Marinelife Aquarium, Mystic, CT 06355.

0 1983 Alan R. Liss, Inc.

94 Overstrom

Fig. 1. Open-mouthed interaction occurring between two dolphins through a barrier net.

would be open and often the jaws touched through the net. Jaw-clapping also was frequently observed. These encounters ordinarily terminated within seconds when one of the dolphins swam away. Because the behavior had components identified previously as sexual or aggressive (see DeFran and Prior [1980] for a review of delphinid behavior), the term “sexual/aggressive response” was used to describe it.

Shortly after the study was completed, the net dividers were removed and the interaction was no longer seen. Recent observations of five bottlenosed dolphins housed in the same facility (January through May 1980) revealed that this behavior, hereafter referred to as the “open-mouthed” interaction or display, was occurring again but in two forms. One form, the head-to-head orientation noted previously, was displayed both between dolphins in the same pool, and between dolphins on opposite sides of a clear plastic barrier separating the holding pool from the main pool (Fig. 2). The second form occurred when one dolphin with jaws open chased and “mouthed” another dolphin in the same pool. This “mouthing” was usually directed toward the caudal region of the fleeing dolphin, yet there was no apparent attempt on the part of the pursuer to bite the pursued. In similar situations Caldwell and Caldwell [1967, p 8951 reported that such sequences often resulted in open-mouthed “attacks.” The purpose of my study was to examine this behavior in more detail.

METHODS

The subjects of the study were five Atlantic bottlenosed dolphins: a male named Kim0 (K) and four females named Cyra (C), Daphne (D), Sassy (Sas), and Seaweed

Behavior of Captive Dolphins 95

MAIN POOL

\

LOWER LEVEL

Fig. 2. Diagram of observation area showing viewing windows, gate, and tape recorder setup.

(S). Cyra, Daphne, and Seaweed were primarily display animals at the time of the study and remained in the 7.3-m diameter holding pool. Kim0 and Sassy were the performing animals of the group and moved from holding pool to main pool through- out the day according to the demonstration schedule. The holding pool is separated from the'main pool by a vertical sliding gate made of clear plastic (0.64-cm thickness) mounted on stainless steel pipe. Dolphins on opposite sides of the gate have visual and acoustical contact when the gate is closed, but limited physical contact.

Behavioral observations were made through large underwater windows in the public viewing area on the lower level. Most observation sessions were initiated prior to the first morning feeding as this was a time of increased, yet undisturbed dolphin activity. Other observations were done randomly, both before and after public hours (9:oO AM-6:oO PM). A checklist technique, as described by Myers and Overstrom [1978], was employed as a quick and easy method of recording behavioral data. Specific behaviors were listed in sequence on a standard data sheet with both the social interactions and the participants being recorded. In addition to the open- mouthed interactions, other social behaviors such as swimming associations, chases, tailslaps, and breaches where one dolphin made contact with another were noted for analyses of social relationships. Jaw-clapping was quantified separately.

Because Kimo and Sassy were regularly let in and out of the holding pool for demonstrations, several social situations could have existed during the study period:

96 Overstrom

all five dolphins together, Kimo or Sassy in the main pool alone, or both Kimo and Sassy in the main pool with the three others in the holding pool. Observations therefore concentrated on animals occupying the holding pool. Only when Kimo or Sassy were seen at the gate, or when they were in with the display dolphins, was their behavior recorded.

In this situation, recording dolphin sounds with a hydrophone while concur- rently observing behavior on the lower level is difficult. Therefore a simple procedure was used to record vocalizations for comparison with the behavioral data. Prior to the study it was noticed that whistles and pulsed sounds produced underwater by the dolphins penetrated the concrete walls of the pools very well. A parabolic reflector (73-cm diameter) was used to concentrate the sounds into a Uher 4400 tape recorder with standard microphone (Model M-17). Extraneous noises were reduced by isolat- ing the microphone in a storage room on the lower level where the main and holding pools met (Fig. 2). The quality of the recordings was obviously inferior to that which could have been obtained using a hydrophone. The procedure was adequate, however, for differentiating whistle types and distinguishing burst-pulse emissions from echo- locative-type pulse trains of slower repetition rate (number of pulses per unit time; Fig. 3). An Ithaco Model 441 filter was used during review of the study recordings to reduce the ambient noise from pool filtration pumps.

Recorder and observer were situated approximately 4 m away from the micro- phone where the gate and dolphins could be seen clearly. Behavioral observations were coordinated with the recordings by noting the metering number from the tape recorder on the data sheet at regular intervals. Jaw-clap behavior also helped to synchronize behavioral and acoustical data.

RESULTS

Thirty-three observation sessions with sound recordings were completed from March 11 to May 6, 1980. Mean duration of each session was 12.7 min for a total study time of 419.1 min. During 78.8% of the sessions Cyra, Daphne, and Seaweed were located in the holding pool with Kimo and Sassy in the main pool (situation I). All five animals were together in the holding pool during 15.1 7% of the sessions (situation 11). Sassy was alone in the main pool while Kimo was with the other females in the holding pool during 6.1 % of the sessions (situation 111).

Over the total observation period 116 open-mouthed interactions were seen. Of these 30.occurred with the participants in head-to-head orientation (mouth-to mouth, Table 1) and 86 with one dolphin directing the action toward the flank of a fleeing dolphin (mouth-to-body, Table 2). Additional social behaviors were described as either nonaggressive (Table 3) or aggressive (Table 4). Jaw-clapping occurred with the following frequencies: Cyra 8, Daphne 25, Kimo 21, Sassy 0, and Seaweed 29.

Pulsed sounds were the only type directly associated with open-mouthed inter- actions. A total of 63 burst-pulse emissions were recorded. These seemed to occur in two categories: short bursts, which were less than 1 s in duration, and long bursts, which were greater than 1 s. As a measure of how long each vocalization lasted, emissions were described by the number of long and short bursts that composed them (Fig. 4). Seven of the recorded emissions had no associated behavior and were determined to have originated out of sight of the observer. In addition, one emission was produced while Sassy was apparently directing an open-mouthed display across

Behavior of Captive Dolphins 97

Fig. 3 . Spectrograms displaying an echolocative-type pulse train (top) and a burst-pulse emission from an Atlantic bottlenosed dolphin. These sounds were recorded prior to the study with a USRD Type H- 23 hydrophone and Uher 4400 tape recorder, and then analyzed with a Kay Electric Company “Missi- lyzer” sound spectrograph using a filter band width of 60 Hz.

the gate toward one of the two belukha whales (Delphinapterus leucas) held in the main pool. Finally, one short burst was recorded during situation 111 that could not be correlated with any particular activity on the data sheets.

Although burst-pulse emissions were not heard during every open-mouthed interaction, there was a higher percentage of burst-pulse emissions heard during mouth-to-mouth interactions (90.0%) than during mouth-to-body interactions (3 1.4 %). Jaw-claps were also more likely to occur with burst-pulse emissions in mouth-to- mouth interactions (51.9%) than in mouth-to-body interactions (37.0%). As the duration of these burst-pulse emissions increased, the occurrence of jaw-claps in- creased as well (Fig. 4). Except for three occasions when jaw-claps were produced unaccompanied by other sounds, the only other vocalizations heard during the inter- actions were pulse trains of slower sounding repetition rate. Jaw-clapping was seen, however, less often with these sounds (13.2%) than with burst-pulse emissions (37.7%).

98 Overstrom

Initiated by: C D K

Sas S

TABLE 1. The frequency of mouth-to-mouth (head-to-head) interactions observed under the three experimental situations (see text for explanation)

Situation I

- 0 12 1 1 14 0 - 3 1 0 4 1 0 0 2 3 1 1 1 0 3

4 3 0 1 0

-

- -

Initiated by: C D K

Sas S

- 0 0 0 0 0 1 - 0 0 0 1 0 0 - 0 0 0 0 0 0 0 0

0 0 0 0 0 -

-

Initiated by: C D K

Sas S

TABLE 2. The frequency of mouth-to-body interactions observed under the three experimental situations (see text for explanation)

Situation I

- 0 0 0 0 0 0 - 0 0 0 0 0 0 1 0 1 0 0 0 0 0

0 0 0 0 0

-

- -

Directed to: Initiated by: C

D K

Sas S

Initiated by: C D K

Sas S

C D

1 0 0 0 1 6 0 0 0 6 0 0 0 0 0 0 1 0 0 1 0 0 0 3

-

- -

- 3 -

5 -

- 14

Initiated by: C D K

Sas S

- - 1 0

0 0 0 0 2 0 2 0 0 0 0 0

0 0 0 0

- - - - - - - - - - - -

- -

Behavior of Captive Dolphins 99

TABLE 3. Nonaggressive behavior; the frequency in which various associations of dolphins were observed swimming together

Situation

Dolphins 1(26 sessions) I1 ( 5 sessions) 111 (2 sessions)

C D K

Sas S

CID CJK

C/Sas CIS DIK

DISas DIS

KISas KIS

SasIS C/DIK

CJDJSas ClDlS

ClKISas CIKIS

CISaslS DlKISas

DIKIS DISaslS KISaslS

CIDIKISas CJKISasIS CIDISasIS

CIDJKIS DlKISaslS

C/DlK/Sas/S

32 1 185

- 218 -

15 17 2 6

10 4

13 1 0 0

13 12 I 2 3 0 0 2 9 0 0 0 0 1 2 0 2 0 0 0

109

14 20 I 1

20 0

13

0 0

-

-

Total 1,106 230 165

DISCUSSION

An examination of the major components of the open-mouthed interaction as well as the chase activity described by Caldwell and Caldwell [1967] suggests an agonistic function for this display. Jaw-clapping is one of the most frequently docu- mented behaviors representing both threat and aggression [McBride and Hebb, 1948; Tavolga and Essapian, 1957; Tavolga, 1966; Caldwell and Caldwell, 19771. More- over, the mouth gape of the bottlenosed dolphin may have threat significance in itself [Caldwell and Caldwell, 19771. The relative frequency of open-mouthed interactions with aggressive and nonaggressive behavior during this study reinforces this conclu- sion. Dolphins that swam together peacefully around the pool displayed a low number of open-mouthed interactions. Dolphins that often interacted by chasing, tailslapping, or making violent contact with one another participated in many open-mouthed interactions (Cyra once directed burst-pulse emissions through the gate to Kimo

100 Overstrom

Initiated by: c D K

Sas S

7 71 1 38

- - 64 -

- - - 31 - - - - - -

- - - - - - 29 - - - 27 2

Initiated by: C D K

Sas S

Situation 111

- 3 0 5 0 8 10 - 1 0 0 11 4 3 - 5 11 23 1 1 1 0 3

1 0 0 0 1 -

-

Directed to: Initiated by: C

D K

Sas S

C D K Sas S Total

4 0 6

0 0 0

2 0 0 2 0 5 0 6

0

- -

- 1 - - -

- - - - - -

- -

100, ,100

1-2~hor1 1 long 2 long 2 2 long bursts burst bursts' bursts'*

emiss ion duration

* Includes 1 long w i t h 1 s h o r t

** Includes up t o 11 long

Fig. 4. The percentage of burst-pulse emissions accompanied by jaw-claps plotted against the relative duration of the emissions.

Behavior of Captive Dolphins 101

during a head-to-head interaction. Kimo made no noticeable response. This one exception was inconsistent with the behavior of the other animals, considering Kim0 and Cyra often swam peacefully together when in the same pool). In addition, postures thought to be associated with submissive behavior, such as a closed mouth and flank area presentation [Caldwell and Caldwell, 19771, were often displayed by a dolphin being chased following an open-mouth interaction with its pursuer. Once this posture was observed in one of the participants of a head-to-head interaction through the gate.

Other evidence lends credence to an agonistic role for the open-mouthed display. Jaw-claps occurred three times more often during open-mouthed interactions with burst-pulse emissions than during those with pulse trains of slower sounding repetition rate. They were also heard more consistently during interactions with long than short burst-pulse emissions. Thus aggression increased with the production and increased duration of burst-pulse emissions. This escalation seemed to occur in a manner similar to that shown in Figure 5 . Interactions often started with one of the participants producing slower sounding pulse trains than burst-pulse emissions. This stage was usually followed by open-mouthed threat displays, burst-pulse emissions, and then jaw-claps if the displays and vocalizations were returned by the other dolphin. Such graded trends are characteristic of aggressive displays in other animals (see Wilson [ 19751 for a review).

The question remains whether burst-pulse sounds are factors contributing to the observed escalation of aggression, or if they simply are the result of heightened motivational levels. Unlike the whistles of bottlenosed dolphins, which are emitted almost omnidirectionally [Evans, 19731, echolocative pulses emanate with highest intensity forward from the melon in a cone of relatively directional sound [Norris et al, 1961; Evans, 1973; Au et al, 19781. Because pulsed sounds of different repetition rates appear to be produced by the same mechanism [Mackay and Liaw, 19811, the projection of burst-pulse sounds is assumed to have the same directional characteris- tics, providing frequency distribution is similar. This means that dolphins in head-to- head orientation project burst-pulse emissions directly at one another. Under condi- tions of high ambient noise, Au et a1 [ 19741 recorded sound-pressure levels associated with the echolocative pulses of bottlenosed dolphins and noted that the average level of 220 dB re 1 @Pa at a distance of 1 m was much greater than previously reported. Sound reception is extremely sensitive in bottlenosed dolphins [Renaud and Popper, 19751 and it is probable that the projection of burst-pulse emissions by one dolphin directly at another could cause auditory discomfort. It is also plausible that tactile discomfort could occur. Norris [I9741 reported studies in which a fellow researcher could actually feel sound pulses on his hand when it was submerged during intense echolocation by a blindfolded dolphin. Hult [1982] and trainers at Mystic Marinelife Aquarium report divers can also €eel pulses when they swim with the dolphins. Thus it seems likely that dolphins can feel the pulsed emissions of conspecifics. Moreover, it has recently been reported that dolphins can use pulsed sounds to herd fish [Hult, 19821 and possibly to stun prey [Norris and Mohl, 1981; Norris, in press]. If the latter is true, then tactile or auditory irritation from burst-pulse sounds at short distances is also a possibility. This could explain, at least in part, the observed escalation of aggression described in this study.

102 Overstrom

INITIATOR RESPONDER

Rapid Pulses Subtle Cues, i.e.,

1

Open Mouth Threat, Burst-Pulses Emitted 9 3

Open Mouth, 3 -< Burst- Pulses,

Vertical Head Movements

5

J A W C L A P pl+

7

Closed Mouth, Quiet

Closed Mouth, Rapid Pulses Returned

2

Open Mouth Threat,

Burst-Pulses Returned

4

Open Mouth,

Bur st- Pulses,

Vertical Head Movemen t s

6

8 Submissive Dolphin Retreats

Fig. 5 . escalation of aggressive behavior.

A hypothetical open-mouthed interaction between two dolphins showing .the proposed graded

CONCLUSIONS

The present study used jaw-claps as an index of aggression during open-mouthed interactions in which burst-pulse emissions were frequently produced. The data and observations suggest that (1) the open-mouthed interactions as described are agonistic confrontations between dolphins, (2) the level of aggressive response between the participants of the interaction increases with the production and subsequent duration

Behavior of Captive Dolphins 103

of burst-pulse emissions, and (3) this escalation of aggression is perhaps related to auditory or tactile discomfort from the reception of intense burst-pulse emissions. Dominance hierarchies have been observed in captive bottlenosed dolphins, and these behaviors may be significant in the establishment of social order among group members.

ACKNOWLEDGMENTS

I am indebted to Dr. William Barry for his invaluable assistance throughout the study. I am also grateful to the Training Department personnel at Mystic Marinelife Aquarium for their extensive cooperation. Sound equipment and technical advice for preliminary investigations were provided by the Naval Underwater Systems Center in New London, Connecticut. Finally, I would like to thank Mr. Stephen Spotte and Dr. William Barry for reviewing and commenting on the manuscript.

This paper is contribution No. 41, Sea Research Foundation, Inc.

REFERENCES

Au, W.W.L.; Floyd, R.W.; Penner, R.H.; Mur- chison, A.E. Measurement of echolocation sig- nals of the Atlantic bottlenosed dolphin, Tursiops truncatus Montagu, in open waters. JOURNAL OF THE ACOUSTICAL SOCIETY OF AMER- ICA 56(4): 1280-1290, 1974.

Au, W. W.L.; Floyd, R.W. ; Haun, J.E. Propaga- tion of Atlantic bottlenosed dolphin echolocation signals. JOURNAL OF THE ACOUSTICAL

Caldwell, D.K.; Caldwell, M.C. Cetaceans, pp. 794-808 in HOW ANIMALS COMMUNI- CATE. T.A. Sebeok, ed. Bloomington, Indiana, Indiana University Press, 1977.

Caldwell, M.C.; Caldwell, D.K. Intraspecific transfer of information via the pulsed sound in captive odontocete cetaceans, pp. 879-936 in SYSTEMS SONAR ANIMAUX, BIOLOGIE AND BIONIQUE. R.-G. Busnel, ed. Jovy-en- Josas, France: Laboratoire de Physiologie Acoustique, Vol. 2, 1967.

Caldwell, M.C.; Caldwell, D.K. Behavior of ma- rine mammals, pp. 419-465 in MAMMALS OF THE SEA; BIOLOGY AND MEDICINE. S.H. Ridgway, ed. Springfield, Illinois, Charles C. Thomas, 1972.

DeFran, R.H.; Pryor, K. The behavior and train- ing of cetaceans in captivity, pp. 319-362 in CE- TACEAN BEHAVIOR; MECHANISMS AND FUNCTIONS. L.M. Herman, ed. New York, J. Wiley and Sons, 1980.

Evans, W.E. Echolocation by marine delphinids and one species of fresh-water dolphin. JOUR- NAL OF THE ACOUSTICAL SOCIETY OF

Hult, R.W. Another function of echolocation for bottlenosed dolphins (Tursiops truncatus). CE-

Mackay, R.S.; Liaw, H.M. Dolphin vocalization

SOCIETY OF AMERICA 64(2):411-412, 1978.

AMERICA 54(1):191-199, 1973.

TOLOGY 47: 1-7, 1982.

mechanisms. SCIENCE 212:676-678, 1981.

McBride, A.F.; Hebb, D.O. Behavior of the cap- tive bottlenosed dolphin, Tursiops truncatus. JOURNAL OF COMPARATIVE PHYSIOLOG- ICAL PSYCHOLOGY 41: 111-123, 1948.

Myers, W.A.; Overstrom, N.A. The role of daily observation in the husbandry of captive dolphins (Tursiops truncatus). CETOLOGY 29: 1-7, 1978.

Norris, K.S. THE PORPOISE WATCHER. New York, W.W. Norton, 1974.

Norris, K.S.; Mohl, B. Can odontocetes stun prey with sound? FOURTH BIENNIAL CONFER-

MALS, Dec. 14-18, Program Section Abstracts, 198 1.

Norris, K.S. ; Prescott, J.H. ; Asa-Dorian, P.V. ; Perkins, P. An experimental demonstration of echolocation behavior in the porpoise, Tursiops truncatus (montagu) . BIOLOGICAL BULLE-

Popper, A.N. Sound emission and detection in delphinids, pp. 1-52 in CETACEAN BEHAV- IOR; MECHANISMS AND FUNCTIONS. L.M. Herman, ed. New York, J, Wiley and Sons, 1980.

Renaud, D.L.; Popper, A.N. Sound localization by the bottlenosed porpoise Tursiops truncatus. JOURNAL OF EXPERIMENTAL BIOLOGY 69:569-585, 1975.

Tavolga, M.C. Behavior of the bottlenosed dol- phin (Tursiops truncatus): Social interactions in a captive colony, pp. 718-730 in WHALES, DOL- PHINS AND PORPOISES. K.S. Norris, ed. Berkeley, California, University of California Press, 1966.

Tavolga, M.C.; Essapian, F.S. The behavior of the bottlenosed dolphin (Tursiops truncatus): Mating, pregnancy, parturition and mother-infant behavior. ZOOLOGICA 42:ll-31, 1957.

Wilson, E.O. SOCIOBIOLOGY; THE NEW SYNTHESIS. Cambridge, Massachusetts, Har- vard University Press, 1975.

ENCE ON BIOLOGY OF MARINE MAM-

TIN 120:163-176, 1961.