This item is only available as the following downloads:
PHONATION PRODU C TION AND SYNCHRONY IN THE BOTTLENOSE DOLPHIN ( T URSIOPS TRUNCATUS ) BY CAITLIN M. O'BRIEN A Thesis Submitted to the Divisions of Social Sciences New College of Florida i n partial fulfillment of the requirements for the degree Bachelor of Arts in Psychology Under the sponsorship of Heidi E. Harley Sarasota, Florida May 2009
! "" To Calvin, Khyber, Malabar, and Rainer t his would not have been possible without you.
! """ Acknowledgements I would like to thank Heidi Harley for sponsoring this research, for being there for the countless hours of editing, and for keeping me hopeful and excited about my thesis throughout the process. Wendi Fellner, for all the time spent coding video data, for answering all of my questions (Dear Wendi, so, the dolphins are making strange sounds a gain), and for teaching how to do things in Excel I never thought possible. Jenna Clark, for constant patience and helpfulness with every spectrogram or statistics question I could think of. Joshua Abbott, for creating amazing codes to analyze my result s in a tenth of the time it would have taken otherwise. Gordon Bauer and Charlene Callahan, for acting as patient committee members through all of the last minute analyses. And Sam the Golden Retriever, for running into Dolphin Lab and sitting with me whenever I needed some reassurance.
! "# Table of Contents Dedication .. ii Acknowledgements .. iii Table of Contents .. iv List of Tables and Figures ... .. v Abstract .. vi Introduction & Literature Review . 1 Method 26 Results 29 Discussion ... .. 32 References .. .. 36
! List of Tables and Figures Tank Set Up26 Phonation Types..28 Participation in Bouts of Synchrony29 Phonation Production Rates30 Whistle Contours ..31
vi vi PHONATION PRODU C TION AND SYNCHRONY IN THE BOTTLENOSE DOLPHIN ( T URSIOPS TRUNCATUS ) Caitlin O'Brien New College of Florida, 2009 ABSTRACT Bottlenose dolphins ( Tursiops truncatus ) have a complex system of acoustic communication that includes several broadly constructed categories of phonati ons: echolocation clicks, whistles, burst pulses, and jaw claps. The structure of this system is still largely mysterious to humans. Social organization among these large brained mammals is also complex For example, calves' period of dependency on their m others is protracted, and adults maintain varied strength associations with many individuals across a wide community. Understanding these multifaceted relationships is of ongoing interest. One measure of social relationships in the dolphin is synchronous b ehavior. Although never previously studied, an understanding of the way in which phonations are related to synchrony might reveal insights both into dolphin acoustic communication and into social organization. To that end, the current study investigated ph onation production rates of 4 male captive dolphins in relation to synchrony. Analysis included 11 five minute simultaneously recorded audio and video files collected throughout a day. Sound files were coded for phonation type and producer(s) when possib le; videos were coded for synchronous swimming The analyses revealed that synchrony was common (28% sampling time) and individual participation levels varied. For example, 1 of the 4 dolphins
vii vii participated in 90% of the synchronous bouts, another in fewer than 30%. Burst pulses and jaw claps, commonly correlated with aggression, were never produced by synchronous dolphins. In files that contained some synchrony, whistles and clicks occurred at lower rates during synchrony than during periods of non synchron y However, rates were lowest in base rate samples in which no synchrony ever occurred Further analyses investigated the use of specific whistle frequency contours including the dolphins' signature whistles, contours typically classified at cohesion calls Whistles were not reliable predictors of synchrony. Overall, phonations did vary in relation to synchrony suggesting that this line of research if worth pursuing in order to learn more about how phonations and social behaviors interact. Heidi E. Harley Division of Social Sciences
Phonation rates and synchrony 1 Bottlenose dolphins ( Tursiops truncatus ) are large brained mammals with complex social structures and a long period of maternal care. They can be found in temperate and tropical waters worldwide. Because of their broad coastal distribution and good health in captivity, they are one of the most well stud ied species of cetaceans Behavioral r esearch on these animals frequently focuses on their aco ustic communication Unfortunately, pragmatic concerns often prevent the study of acoustic communication in concert with observed behavior. Pas t research suggests that some types of vocalizations may be correlated with certain behavioral contexts, such as signature whistles with physical isolation from conspecifics (e.g., Janik & Slater, 1998) To date, several studi es have investigated synchrony the coordination of resting, travelling, breathing, and other motions, in bottlenose dolphins; thus far, none have combined these observations with acoustic analysis The refore, the current study investigates how the phonations, or sounds of both vocal a nd physical nature, are related to synchrony. The Social Dolphin Daily Life Dolphins are social animals and are often seen in groups. Group size depends on a variety of factors, including danger of predation and availability of food (Reynolds, Wells, & E ide, 2000). Atlantic bottlenose dolphins often live in smaller social groups in shallow, co a stal areas because food is easy to find and risk of predation is relatively low. However, hundreds of dolphins may group together in the deeper oceanic waters since food distribution tends to be patchier and risk of predation is higher. Typically, individuals of the same age and reproductive status group together. For example, researchers have noted
Phonation rates and synchrony 2 that some common group formations include females or females and the ir calves, subadult dolphins, or male pairs. Though location has some impact on the daily activities of dolphins, most groups spend the majority of the day traveling (Reynolds et al. 2000). Feeding and socializing are two other frequent activities. Feed ing behaviors are one illustration of this species' plasticity; often, dolphins in different regions will pursue food in special ways. Some dolphins intentionally strand themselves and feed on fish they have herded to shore (Hoese 1971, Silber 1995), whi le others follow fishing trawlers, feeding on discarded by catch (Fertl & Leatherwood, 1997). Herding (Gazda, Connor, Edgar, & Cox 2005), fishwhacking, in which dolphins smack fish out of the water with their tail flukes (Wells, Scott, & Irvine 1987), and burrowing after fish buried in the sand, known as crater feeding (Rossbach & Herzing 1997) have also been observed in different wild populations. Dolphins also spend a portion of the day engaged in play. Kuczaj, Makecha, Trone, Paulos and Ramos (2006) no te that "play appears purposeless (at least in terms of immediate consequences)," and that "play may be solitary or social," with social animals like bottlenose dolphins engaging in both types of play. Formation and Function of Alliances Male bottlenose dolphins are often seen in relatively stable pairs called dyads. In some cases, the dolphins may also form groups of three, called triplets (Connor, Smolker, & Richards, 1992). These pairings can last several months or more than a decade (Connor, Mann, & Read, 2000). Parsons, Durban, Claridge, Balcomb, Noble, & Thompson (2003) studied a wild population of male bottlenose dolphins in the Bahamas to investigate the composition of these dyads. They hypothesized that male dolphins form alliances with
Phonation rates and synchrony 3 more clos ely related kin, as has been seen in several terrestrial mammal species. Collection of skin tissue and fecal samples allowed the researchers to determine the relatedness of the observed dyads. In all, 11 dyadic alliances were identified in the study. The r esearchers were able to get samples from both males in six of these alliances. By analyzing a section of the dolphins' mitochondrial DNA, the researchers determined that these male dolphins tended to form dyads with maternally related individuals. However, data from Sarasota Bay provide conflicting results (Duffield & Wells 2002), and it seems probable that dolphins in different locations follow different affiliation practices Connor et al. (1992) found that alliances can exist on at least two levels, and that these may have slightly different functions. First order alliances like those described by Parsons and colleagues are more closely knit, usually consisting of two or three males but dolphins can also form larger, more fluid groups In Connor's sa mple in Shark Bay, Australia, 21 males were tracked over 25 months from 1987 1989 through collection of survey data. Association coefficients were calculated to determine how often two dolphins were sighted together, and data revealed that some alliances w ere sighted together as frequently as mother dolphins and their calves In Shark Bay, m ale dyads or triplets often work together to herd female dolphins aggressively. Connor focused on one triplet alliance (Snubnose, Bibi, and Sicklefin) in particular and found that, over the course of one year, the alliance herded 48 dolphins. Thirty of these were confirmed females, and the researchers assume the other 18 were also female. Researchers did not observe the beginning stages of all herding events. However, in those they witnessed, they noted, "male displays often included striking synchronous underwater turns and aerial leaps" (Connor 1992 p. 987 ). Interestingly, only two of the three males would be involved in any herding
Phonation rates and synchrony 4 ev ent, while a third dolphin was odd male out The composition of the her ding pair and the odd male out varied over herding events, and occasionally the odd male out would join the males of a different alliance for a short time. Connor and colleagues (1992) also found that each first order alliance associated frequently with one or two other first order alliances. While a small alliance seems to be successful for female herding in Shark Bay, larger groups were created when stealing females from other first order alliances. For example, pair A and triplet B joined together to steal females from triplet H in four observed instances. Twice, triplet D assisted triplet H in response to alliance A and alliance B's theft attempts. When two alliances joined together to steal females from lone first o rder alliances, the 2 vs. 1 competition (n = 4) resulted in quick resolution (<2 min.). However, when the defending' alliances were joined by yet another assisting alliance (2 vs. 2 competition), the competition between groups lasted much longer. One of t he two observed instances lasted more than 10 minutes, while the other dragged on for more than an hour. While larger alliances seem to increase access to females in many cases, formation of second order alliances can also intensify competition. To test th e hypothesis that herding behavior might serve a reproductive function, the researchers compared the pregnancy statuses of females herded by the Snubnose, Bibi, and Sicklefin in 1987 and 1988. The data show that the males were more likely to herd nonpregna nt females compared to pregnant females supporting this hypothesis. While dyads and trios are common, not all males appear to form stable alliances with one or two other dolphins. Connor, Heithaus, and Barre (2001) found that male bottlenose dolphins in the Shark Bay population sometimes form a super alliance, defined
Phonation rates and synchrony 5 as "a very large second order alliance in which members jointly attack other alliances (p. 263). In contrast to the stable dyads and trios studied in 1992, trio formation within a super all iance is flexible. Males often switch trio partners between consortships with female dolphins. Synchrony Synchronous swimming is seen frequently among wild and captive dolphins. Allied males and mother calf pairs engage in synchrony most frequently, though synchrony is sometimes noted between other age and gender combinations. Synchrony has been defined as "simultaneous movement in a parallel orientation" (Fellner Stamper, Losch, Dahood, & Bauer, in prep. ) but can include resting behavior as well. Sy nchronous swimming seems to have many benefits. It may reduce the risk of predation for all dolphins, but especially for calves. Additionally, synchrony allows a calf to be caught in its mother's slipstream, reducing energy costs to keep up with the mothe r (Noren, Biedenbach, Redfern, & Edwards, 2008). Fellner (2008) investigated the ontogeny of synchrony between a captive bottlenose dolphin and her calf. The two were housed at the Seas, Epcot. Video recording began immediately after the calf's birth and c ontinued for five weeks. For the 12 hours following the birth of the calf, recording was constant. After that, the dolphins were recorded for 5 minutes every two hours. On average, the mother calf pair spent 97.58% of their time in synchrony. Fellner also used Hinde's Index of Association (Hinde & Atkinson, 1970) to determine if either dolphin was primarily responsible for maintaining synchrony. This index is calculated by comparing the relative number of approaches to the departures of pair members Fellner found that the mother
Phonation rates and synchrony 6 was primarily responsible for maintaining synchrony, and that she was quick to respond when her calf broke synchrony ( M = 0.759 seconds). Two additional females joined the mother calf pair one week after the calf's birth. Fe llner examined synchrony between the calf and these two females as well as the development of ten different behaviors in the calf: turn on side, belly up, roll, tail slap, spit, chuff, porpoise, stop, object toss, and bow. One female, Nina, swam synchronou sly with the mother calf pair much more frequently (55% of the time) than the other, Snapper (9%) However, the mother was involved in all bouts of synchrony between the calf and the other females. The data on behavior development reve aled that the calf mo deled the listed behaviors after the adults' actions. In all but one case, one of the adults performed the behavior before the calf. In addition to modeling, synchrony seemed to play an important role in behavioral performance ; 81.7% of the 514 bouts of th e behaviors were performed while swimming in synchrony with an adult. When adults engage in synchronous behavior, it seems to serve an affiliative purpose (Connor, Mann, & Watson Capps, 2006). In the Shark Bay population, female dolphins engage in contac t swimming, in which "one dolphin (actor) rests its pectoral fin against the flank of another dolphin behind the other dolphin's pectoral fin and below or just posterior to the dorsal fin" (Connor et al. 2006). Connor et al. recorded a ll instances of cont act swim ming in a focal dolphin's group To determine duration, the researchers counted a new case of contact swimming if five minutes or more had passed since the previous contact swimming event, if one individual broke off a contact swim to swim with ano ther individual, or if the individuals switched roles. The researchers found that contact swims overwhelmingly occurred between adult female pairs (83.3%). In contrast, only
Phonation rates and synchrony 7 seven (7.3%) of the events occurred between two males, and only one of these was b etween two adult males. Female male contact swims were also rare (6.3%), and all of these involved juvenile males. Average contact swim time was less than five minutes, thoug h one adult female pair contact swam for 34 minutes. The results revealed that fem ales were most likely to contact swim with other females in male biased groups. The researchers hypothesize d three possible be nefits of female female contact swimming in these male biased groups; it may reduce male harassment, assist in locomotion, or redu ce stress levels. The prevalence of this female female contact swimming is interesting in part because females typically form much weaker associations than males. In another investigation of multi sex groups, Perelberg and Schuster (2008) researched the links between breathing coordination, association patterns, and dyadic spatial formation. Thirteen dolphins were observed in this study. The animals were housed in a semifree enclosure in the Red Sea, south of Eilat, Israel. The group consisted of 5 males and 8 females. One male and 4 females were adults, two males and two females were adolescents, and two females and two males were calves. The dolphins were free to come and go from the enclosure to the open sea via a gate in the net barrier. To measure ass ociation patterns, the researchers observed dolphins during the 20 minutes before the four daily feeding sessions. The location of each dolphin was noted every minute, and their position s were classified in one of four categories: to the left of the feeding pier (L), to the right of the feeding pier (R), away from the pier (A), or outside the enclosure area (O). Associations were calculated using the Half Weight In dex, which yields a score from 0 (no association) to 1 (tight association). To determine if the se associations were based on attraction, avoidance, or random variations the researchers used Monte
Phonation rates and synchrony 8 Carlo procedure iterations to generate a frequency distribution for each dyad. If the value exceeded 0.025 probability on the high or low tails, the assoc iation was considered significant. Breathing coordination was classified into one of four levels: (a) if breathing was simultaneous, (b) if the second dolphin came up to breathe while the first was at the surface, (c) if the second breathed after first wa s submerged, or (d) if more than 10s elapsed between the breathing incidents of the two dolphins. The mean breathing coordination level was calculated for each observed dyad for each session. To assess dyadic spatial formation, researchers recorded the lo cations of each member of the pair during breathing coordination. Spatial formation was divided into five categories: (F) if subject dolphin was in front of partner, up to one bo dy length apart; (A) if the subject was ahead of its partner, with over lap of up to half a body length; (P) if both dolphins were parallel up to h alf a body length in difference; (L) if the subject lagged behind its partner with an overlap of up to half a body length; or (B) if the subject was behind his partner, up to one body len gth apart. "F ront "& "Behind" and "Ahead & L agging" were considered equivalent for purposes of this experiment. The results showed that age and gender had a significant effect on association patterns, with mother/calf pairs (0.7 HWI) and the adolescent ma les (0.8 HWI) showing highest levels of attraction. Mother/calf pairs and same sex adolescent pairs showed higher than random coordination while breathing, while lower than random coordinati on was found in the adult male adolescent female pairs. The adult male tended to take the "Ahead" or "B ehind" pos ition with adult females, the "A head" position w ith adolescent males, and the "Ahead" or "L agging" positions with adolescent females. Adult females did
Phonation rates and synchrony 9 not show an overall trend while swimming with each other, but tended to take the "Front" or "A head" positio ns with adolescent males, the "A head" position wit h adolescent females, and the "A head" position with their calves. Adolesc ent males tended to be in the "P arallel" position while s wimming together, and to t ake "L agging or "P arallel" positions with adolescent females. Adolesce nt females tended to take the "Ahead" or "L agging" positions while swimming together. Mother calf pairs and the adolescent male dyad showed an expected link between strong levels of attraction and good coordination while breathing. However, the adolescent female pairs showed high breathing coordination while showing a weak association. While adult male adult female pairs tended to be in the same areas of the enclosure, the breathing c oordination of these pairs was highly variable, and they showed distant spatial position. Since coordinated breathing is usually associated with strong levels of attraction, Perelberg and Schuster suggest the action is a form of cooperation, and that it is influenced by both the immediate material outcomes and by the social bonds formed when the animals coordinate behavior for j ointly obtained outcomes. T he results of this paper are also important because they suggest that male pairs are most likely to swim in a parallel position, which is one of the clearest displays of synchrony. Sometimes synchrony can be observed in post conflict situations. Tamaki, Morisaka, and Taki (2006) observed three captive bottlenose dolphins for aggressive interactions and flip per rubbing to determine if post conflict flipper rubbing is correlated with a decrease the likelihood of another attack, and to determine rates of post conflict flipper rubbing between the two opponents and between either opponent and a third bystander do lphin. Two dolphins were mature females, while the third was an unrelated
Phonation rates and synchrony 10 male calf (3 years of age). The dolphins were observed between 9:00 a.m. and 8:00 p.m. for 23 days between May and October, 2003. Though observation times varied from day to day, obs ervations of less than 30 minutes were excluded from the analyses. In addition, no observations were conducted during feeding or performance times. The researchers classified the following behaviors as aggressive: threatening, chasing, biting, hitting, bo dy slamming, and head butting. So as not to include interactions that were actually play, the researchers only noted interactions in which the receiver of the action attempted to evade the initiator via fleeing or flinching, as defined by Samuels and Giffo rd (1997). In addition, chasing was only considered as aggressive if it was accompanied by burst pulse vocalizations or if no role changes occurred, since previous research suggests that chasing with turn taking is actually play (Mann & Smuts, 1999). Beca use the recipient dolphin often retaliated, an end to an aggressive encounter was defined by a lapse of more than one minute after the cessation of aggressive behavior. Pectoral fin rubbing was classified in a similar way; the incident was ended when the r ubber moved away and ceased rubbing for more than a minute, or when the rubber and rubbee switched roles. Pectoral fin rubbing occur ed more frequently after an aggressive encounter than during control periods, but this effect was only found when the young male was involved in the aggressive encounter. Post conflict flipper rubbing was not seen between the two adult fem ales. Post conflict rubbing was associated with increased time between aggressive behaviors. Also, flipper rubbing between an opponent and a third party was more common after an aggressive interaction than during the control period.
Phonation rates and synchrony 11 These results suggest that dolphins use pectoral fin rubbing to repair relationships after conflict, and it is an effective way to increase the time between aggres sive encounters (Tamaki, et al., 2006) It is important to note that this effect might only hold when a juvenil e is involved in the encounter and that the findings are based on correlations so causation can only be inferred S ince the sample in this study was small, these types of behaviors should be monitored in more diverse captive and wild populations. Dolphin Phonations Dolphins are capable of producing several distinct types of sound: whistles, echolocation clicks, and burst pulses. Acoustic commu nication is the most viable medium over di stances in a marine environment. Sound travels far and fast underwater. Quintana Rizzo, Mann, and Wells (2006) determined that acoustic range is limited by noise in the environment, such as turbulence or bottom com position, rather than hearing sensitivity. In channels, whistles between 3.5kHz and 10 kHz could be detected more than 20km away (Janik 2000). Whistle detection range decreased with increased water turbulence. In contrast, higher frequency whistles (12 kHz) did not travel as far; detection range for these fell to 1.5 4 km. Echolocation clicks can also travel long distances. Au and Snyder (1980 ) found that a male bottlenose dolphin could perceive a 7.62 water filled steel sphere f rom a maximum distance of roughly 113 m. Whistles Dolphins produce a variety of whistles. These are narrowband, frequency modulated sounds that last from about 0.1 to 2 sec. The most studied whistles are individually distinct whistles known as signature w histles. Dolphins develop their own
Phonation rates and synchrony 12 signature whistles as young calves. Signature whistles are important for maintaining cohesion between mothers and calves, as addressed by Smolker, Mann and Smuts (1993). In a period of about 3 months, 114 hours of visu al and acoustic data were recorded during focal follows of dolphin infants. This study utilized data obt ained from 6 calves, and focused primarily on 2 male calves. The signature whistles of each calf had to meet these criteria: it was the predominant whis tle in recordings of the infant, it was the only whistle type occurring when the dolphin was alone, and the whistle could not be present when the infant was not present. Smolker et al. categorized three kinds of mother/calf separations: infant position bre aks, close separations, and far separations. Position breaks were classified as short distance separations in which the calf left infant position to a distance to <5m before returning, or up to 20m if the calf returned in <2 minutes. Close separations were defined by a separation of 5 20m for more than 2 minutes. Far separations were classified as separations of more than 20 m between the pair. Results showed that the calves produced more signature whistles as the distance of separation increased. Additio nally, whistle rates were higher towards the end of the separation period. The researchers suggest that the calves' signature whistles facilitate reunions between the mothers and calves by alerting the mothers of the calves' locations and possibly making t he mothers aware that the calves are attempting to reunite with them. For example, a travelling or foraging mother might slow down to allow her calf to catch up upon hearing its signature whistles. However, the calves were the subjects followed in this stu dy, so it is difficult to say how the calves' whistles influenced the mothers' behavior. Bottlenose dolphins are also able to produce the signature whistle of others In one study, Janik (2000) investigated the occurrence of whistle matching in wild bottleno se
Phonation rates and synchrony 13 dolphins using a noninvasive passive acoustic localization technique. By comparing the sounds recorded by three widely spaced hydrophones (underwater microphones) Janik was able to ensure that whistles were coming from different dolphins. H uman sorters naive to the origins of the whistles categorized spectrograms, graphs of sound frequency and intensity relative to time, in whistle interactions on a scale from 1 5, with 5 being most similar. If a whistle interaction received an average score of 3 or hig her, it was considered to be a matching interaction. An average of 10 animals were present in the channel at any time. Thirty nine of the 176 whistle interactions in this sample were categorized as matching, and further analyses revealed that the distance between animals in matching interactions was much smaller than in non matching interactions. To ensure that repeated whistles from the same dolphin were not included, the analyses may have missed some whistle matching interactions. The do lphins had to be m ore than 26 m apart to be localized by the hydrophone array In seven of the matching interactions, the whistles of the dolphins overlapped. While this mimicry has been related to aggression in birds, it is uncertain what purpose it serves for dolphins. W atwood, Tyack, & Wells (2004) determined that male bottlenose dolphin dyads exhibit whistle convergence. Data were collected from 1985 2001 during health assessments of wild dolphin populations. During these times, individual dolphins were captured and res trained for a short period of time. Using a suction cup hydrophone, the researchers were able to record the vocalizations of each restrained dolphin. A total of 15 males and 2 females were recorded. The males represented 9 distinct alliances, some due to t he re pairing of a male once his original d yad member died or disappeared. Human judges, who were nave to whistles from this dolphin population and did not know the
Phonation rates and synchrony 14 identities of the whistlers or how many whistlers were represented in the sample, rate d th e similarity of 465 pair wise comparisons of whistles made by 5 dyads on a scale of 1 5. M ales shared significantly more whistles with their partners for 7 of the 18 paired males. The researchers conclude d that male bottlenose dolphin dyads display whistle convergence, though its purpose is unclear. It is possible that males exhibit such whistle similarity because many grow up in the same social groups, but some dolphins in this sample that came from different groups show ed high levels of convergence, while other animals were reared in the same groups and show ed hardly any convergence. Therefore, the researchers concluded that the pair bond between two males makes whistle convergence far more likely to occur. Another study of whistles in dolphin groups did not focus specifically on male dyads. Quick and Janik (2008) determined that wild dolphins whistle less when group size greatly increases, quite possibly to avoid masking by other conspecifics. The study was conducted on wild populations in the Moray Firt h, Scotland using passive acoustic localization and focal follows. The passive acoustic localization technique uses four hydr ophones to allow researchers to assign whistles accurately to a specific direction, which may then be correlated to dolphins' surf ace behavior. Quick and Janik observed both a focal group and surrounding groups using 2 minute point sampling. The researcher observing the other groups recorded how far other groups were from the focal group, and another researche r recorded behavior cont inually and recorded group size, swimming direction, and distance from the boat in two minute intervals. A "group" was defined as two or more individuals, and all indi viduals could be, at most, 10 m from another group member. Since visibility was crucial, data were only collected on calm, clear days.
Phonation rates and synchrony 15 The behavior of the focal group fell into one of four categories. Surface travel was characterized by movement in the same direction without creation of white water in the dolphins' wake. Nonpolarized movement was classified by nondirectional movement and surfacing. Socializing was defined by interaction in close proximity, including body rubbing, rollovers, belly displays, and fins and heads sticking out of the water. The final category included jumping and body slaps, defined by aerial behavior and splash ing. A total of 1,783 whistles were analyzed. The data revealed that behavior had a significant effect on whistle rates; for example, whistle rates during nonpolarized movement were significantly higher than those during surface travel ( p < 0.001). Whistl e rates during socializing were also significantly higher ( p < 0.01) than rates during surface travel. Analysis revealed that group size had no significant effect on whistle rate ( p = 0.793), which means that whistling per individual went down as group siz e increased While whistle rates actually increase in frequency for groups up to about 10 animals, they begin to decrease in cases where group size exceeds about 15 animals. Quick and Janik suggest that the increase in whistle frequency during milling vers us surface travel may occur because individuals are easily lost in the group during the unstructured movements of milling. Also, milling may be a transitional state preceding a behavioral change, which may account for increased whistling. It is possible th at the dolphins whistled more during socializing to maintain contact with other dolphins in the group, since social behavior can cause the group to become dispersed. The researchers also suggested that the decrease in individual whistling in very large gro ups might occur to avoid masking, which would render vocalizations ineffective.
Phonation rates and synchrony 16 Echolocation clicks Echolocation plays a very important role in the lives of toothed whales ( Odontoceti ). Echolocation c licks allow for navigation and prey detection even in waters with poor visibility (Madsen, Kerr, & Payne, 2004) Many experiments suggest that dolphins use echolocation clicks for object recognition. For example, Pack Herman, Hoffman Khunt, & Branstetter (2002) carried out an experiment to determine how d olphins perceive echoes. Their study tested the ability of an adult female dolphin, Elele, to match from a sample, which she could see, to three alternatives, which she was only able to echolocate. The objects differed in global shape, but shared a few loc al features. For example, capital and lowercase L's vary in global shape, but they share the local feature of a vertical line. Elele performed well in this task, choosing the correct match 93% of the time Next, they tested the reverse; her ability to matc h an echoic sample to visual alternatives. Elele performed even better on this task with a 99% accuracy rate. In both types of match to sample (MTS), she searched from right to left in a serial, nonexhaustive manner; Elele stopped searching once she encoun tered the correct object. It should be noted that the test subject in this study had previous experience with V E and E V matching and had also been trained in V V and E E matching. In addition, while the object combinations used in this study were novel, the objects themselves were highly familiar to Elele. In a second experiment, the researchers added a fourth obje ct to the array of alternatives: a paddle that Elele could press if none of the alternatives matched the sample. This null' possibility was ad ded to investigate further whether Elele was using local or global cues. If Elele utilized local cues, she would find them in one of the three object alternatives. Even though it would not match the sample in global shape, Elele would use the "null" paddle less often
Phonation rates and synchrony 17 because of the presence of local cues. However, if she were using the global shape of the object, she should use the null paddle more often when the match was not present. As in the first experiment, Elele was tested with V E and E V matches. W hen the correct match was present, she chose it at highly significant levels (94% and 95%). When the correct object was not present in the alternative array, she pressed the null paddle in 74% of V E trials and in 76% of E V trials. The researchers analyze d her errors and conclud ed that she was more likely to choose an object incorrectly on the left than on the right. Since her search pattern was always from right to left, the researchers conclude d that she was more likely to mistake an incorrect object for the correct match when more time had passed in the search, suggesting that errors resulted from a weakening of short term memory. These data loosely support the conclusion that echolocating dolphins perceive objects in terms of global, not just local, fea tures. However, methodological concerns are abundant, so results should be interpreted cautiously. Echolocation clicks may also serve a social function. Xitco and Roitblat (1996) experimentally showed that dolphins are able to obtain object informa tion through eavesdropping on other dolphins echoes. Two adolescent, captive male dolphins housed in central Florida participated in this study. First, both dolphins were trained in a n echoic match to sample (MTS) task with a shared set of 18 objects. The MTS task required a subject to examine one object, the sample, and then choose the matching object from a 3 alternative array containing an object identical to the sample and two distracter items. In this study, the match was done echoically. This means t hat the dolphin was not able to see the sample object when it was presented, but was able to echolocate it through thin, opaque plastic. After examining the sample object, the dolphin swam to a box containing the
Phonation rates and synchrony 18 choice objects. As with the sample, these c hoices were not visible to the dolphin but were echoically available. The dolphin selected an object by pointing to it with his rostrum for 2 s. A preliminary testing phase was conducted with the set of 18 familiar objects. During testing, the dolphin with a lower accuracy rate in the MTS task consistently acted as the object investigator while the other was always the non echolocating l istener. While the investigator echolocated the sample with his head stationed in a hoop underwater, the listener dolphin was stationed on a bite plate with his rostrum in the water but his blowhole and melon out of the water. This allowed the listener to receive the echoes produced by the investigator via the lower jaw, but since his melon was held above water, he was unabl e to examine the object himself. The listener and investigator then chose an object from two separate arrays, each containing the correct match and two distracter objects. The listener chose the correct object significantly more often (50%) than expected b y chance (33%). Additionally, the listener was correct more frequently when the inspector chose the correct object, though this difference was not significant. Xitco and Roitblat (1996) then repeated the experiment with a new set of stimuli. It included 9 objects from the first experiment and 18 new objects. Each dolphin was familiar with a different half of these new objects. Testing procedures followed those of experiment 1. In these tr ials, the listener scored significantly above chance (47%). However, if trials with shared familiar objects are excluded, performance did not exceed chance (31%). Further analyses showed that the accuracy of the inspector had a significant effect on the li stener's choice. When the inspector was correct, the listener's performance was above chance ( 59%). However, an incorrect choice by the inspector often resulted in an incorrect choice by the listener; in this case, he was correct in only 29% of trials. The researchers
Phonation rates and synchrony 19 suggest that this may have resulted from insufficient signal strength on the part of the inspector. Xitco and Roitblat note d that the listening dolphin often slid even closer to the inspector than was required, suggesting that close orientatio ns may allow for best signal reception. These results are important because they provide strong evidence the sharing of information via echolocation. Bu r st pulse s Although much research has been conducted on the narrow band whistles of dolphins, far less h as been done to determine if broadband sounds might also have a social function. Burst pulses are broadband sounds like echolocation clicks, but the interclick interval in burst pulses is very short creating a sound that humans hear as more of a buzz than a click train Overstrom (1983) conducted one of the first studies relating burst pulse sounds to behavior. Five dolphins (1 male, 4 female) were housed at the Mystic Aquarium in Connecticut during the study period. Three of the females were display anima ls and spent the study period in a holding tank. This holding tank was connected to a larger main tank, and a clear plastic divider separated the two. The fourth female and the male were performance animals and were moved between the holding pool and main pool throughout the day. The behavior of the two performance animals was only analyzed when they were in the tank with the display animals or when they were at the clear divider. A total of 419.1 minutes were recorded over the study time and 116 open mouth interactions were observed. Open mouth postures are considered a threat, and in this study, they were seen in mouth to mouth orientations 30 times and in mouth to body orientations 86 times. The researchers reco r ded sound concurrently and found that pulsed sounds were the only phonations directly associated with these open mouth threats. A total of 63 burst pulse
Phonation rates and synchrony 20 emissions were recorded, with most burst pulse emission occurring during the mouth mouth interaction Mouth to body interacti ons included the production of burst pulse sounds less often Overstrom was also interested in the production of jaw claps, loud, percussive sounds created when a dolphin opens and closes its mouth rapidly, and he found that jaw claps were also more freque nt in mouth to mouth orientation, and that increased duration of the burst pulses or burst pulse trains multiple closely spaced burst pulses, was correlated with an increased production in jaw claps. More recently, Blomqvist & Amundin (2004) utilized arc hival video and audio recordings of two female dolphins from a captive colony at Kolmrden Zoo, Sweden (from Karlsson 1997), and recorded new audio and video data on all dolphins in the colony, including the two females from the older recordings plus 10 o ther dolphins. ( Karlsson's 1997 study of the two female dolphins had revealed that burst pulse sounds with pulse repetition rates between 100 900 pps were common during aggressive interactions. ) During recording sessions in Blomqvist & Amu n din's study, th e 12 dolphins were temporarily separated into two groups via a net barrier. A hydrophone was suspended from the barrier approximately 1.5 m below the surface. The researchers classified an interaction as aggressive if two dolphins were face to face on eith er side of the net barrier and emitted burst pulse sounds while producing behaviors like head jerks, pectoral fin jerks, S shaped body postures, and/or jaw claps. The distance between dolphins was generally much greater with the free swimming females (10 20 m.) of Karlsson's study than during interactions across the net barrier (1 4 m.). The dolphins on either side of the net pointed their rostrums towards each other while emitting burst pulse sounds. Often, the aggressive behaviors escalated to a climax of high
Phonation rates and synchrony 21 frequency burst pulse sounds in conjunction with fast jaw claps and other aggressive body postures. If the net was removed during an aggressive interaction, the dolphins would swim quickly and silently around the pool, ceasing aggressive behaviors. Analysis of the burst pulse sounds reveal ed characteristics that set them apart from sonar click buzzes', such as those that are heard when an echolocating dolphin approaches a target object. The highest pulse repetition rate of the burst pulse sounds reached 940 pps. This signifies a pulse interval of a little over 1 ms. T he average distance betwe en dolphins was about 1 4 m which is much greater than the distance seen in close range echolocation signaling The researchers conclude that these high frequency burst pulse sounds have a social function for dolphins. They speculate that such intense burst pulses may cause physical discomfort or pain in the targeted dolphin, and that these sounds may function as a way to avoid more physically damaging altercations. Other phonations While whistles, burst pulses, and click trains are common in almost all recorded dolphins, several other kinds types of vocalizations have been noted in wild populations. These may be cultural or behavior specific phonations, or it is possible that individual researchers classify similar vocalizations differently. Connor and Smolker (1996) found that male bottlenose dolphins in the Monkey Mia shallows and the greater Shark Bay produc e a popping vocalization. Usually, this sound is produced when an alliance of males is flanking a female dolphin, although pops do occur in all male groups. This study focused on three males that frequented Monkey Mia to receive handouts of fish from the l ocals and tourists. Often, the males would have a single female with them, and they remained close to her at all times. In much dolphin
Phonation rates and synchrony 22 vocalization research, it is difficult for the researchers to determine which dolphin is producing the sound. Luckily, t he males often produced the pop' sound above water, enabling the researchers to attribute pops to individuals. Although pops were also produced underwater, Connor and Smolker focused only on the pops that occurred above water. In addition to recording the pops, the researchers also observed the females' behavior in response to the sounds. Specifically, the researchers were interested in incidents when the female changed direction and turned toward the male within 15 seconds of the popping. Therefore, the r esearchers only recorded incidents when the female was originally parallel to or oriented away from the male. The results revealed that typical pop trains included 3 27 pops, usually with 6 12 repetitions per second. The approximate durations of the pops ranged from 4 11 ms though it was difficult to decipher when a pop began and ended from the spectrograms. Interestingly, the researchers noted that the popping male's blowhole seemed to be pushed up on one side during pop production, and that louder pops seemed to correlate with a greater rise in the lips of the blowhole, suggesting that the sounds were made through the blowhole rather than through jaw claps or other behavior. The researchers were able to observe the males with 12 different females in 19 consortships over a period of 31 days. A consortship was a period of time during which the three male dolphins surrounded and controlled the movements of a single female dolphin (usually in estrus, to gain mating access to her, see Connor 2001). Twenty six of the 31 sessions yielded at least 10 samples of both pops and baseline data. In all of these sessions, the females approached the males at higher rates when the males were popping than they did during baseline. In addition, the researchers analyzed the effect of pop train length on
Phonation rates and synchrony 23 female behavior. They found that females were significantly more likely to turn to the male in response to a longer pop train (>5 pops) than during a brief pop train ( 5). The researchers concluded that pops are a threatening social signal used to keep the females close to the popping males. Indeed, previous research (Overstrom, 1983) revealed that popping is often correlated with head jerks, a signal of aggression. In another study, Janik (2000) investigated the use of a bray call produced by wild bottlenose dolphins in the Moray Firth, Scotland. Acoustic recordings were made using a passive localization system. By placing hydrophones at different points in the channel, it was possible to identify individual vocalizers. Janik focused solely on braying in this study and monitored the behavior of other dolphins in response to the first bray. These behavioral observations were conducted from the surface of the water. In particular, Janik looked at fast swimming, leaping, feeding, and surfacing. To determine if the dolphins tended to respond to bray calls by swimming quickly towards the calling dolphin, Janik broke fast swims into three categories. If the fast s wimming dolphin was within 50 m of the calling dolphin, it was said to o ccur at the position of the caller. Fast swimm ing dolphins more than 50 m away were either categorized as swimming towards or away from the calling dolphin. Fast swimming increased significantly both at and towards the position of the caller after a bray, while fast swimming away from the caller was rare. While fast swimming at the location of the caller could be related to chasing prey items, fast swimming towards the caller indicated that conspecifics had been attracted to the area. These results suggest that brays are food related calls that result in the approach of other dolphins. However, since the bray is such a low frequency sound, Janik proposes that the primary purpose of the bray is prey manipulation. The dolphins in the study were feeding on
Phonation rates and synchrony 24 salm on, which can be disoriented by low frequency sound. It is possible that other dolphins respond to this vocalization by approaching so that they, too, have the opportunity to feed. Phonation Production Rates Pressman (2007) sought to establish base line rates of three dolphin vocalization categories: whistles, echolocation clicks, and burst pulse sounds. These baseline rates were tracked over a variety of conditions in which total group size and in group pairings varied. When the study began, the fac ility where the study was conducted housed two male dolphins. Ranier was approximately 25 years old, and Calvin was approximately 12 years old. Two additional dolphins, Khyber (14 y.o.) and Malabar (6 y.o.) were introduced to the facility over the time of the study. The results showed substantial changes in vocalizations over three recording periods: pre during, and post introduction between the two resident males, Ranier and Calvin, and the two new arrivals, Khyber and Malabar. On average, whistles and click trains occurred most commonly during the introduction period, but average production of burst pulse sounds, which are commonly associated with aggression, was markedly higher across post introduction r ecordings. Trainers noted that the youngest male, Malabar, was often pursued aggressively during production of these burst pulse sounds. Pressman concluded that group size alone cannot allow us to predict the frequency of vocalizations. Social dynamics also come into play, such as dolphin pairing s and no vel situations. Current Study As the research reviewed above demonstrates, we know that interaction with conspecifics is important to bottlenose dolphins. The Shark Bay and Sarasota Bay longitudinal studies have shown us that males may bond very tightly, forming pairs or trios
Phonation rates and synchrony 25 that can endure for decades (Connor Heithaus, & Barre 1999 ). Often these pairs rest, leap, dive, and travel in synchrony (Dudzinsk i & Frohoff, 2008). Wild males might benefit from synchrony in several ways. As with mother calf pairs (Noren, Biedenbach, Redfern, & Edwards 2008), synchrony between males may provide a hydrodynamic advantage. It may also serve as a social cue that the synchronous males are "a solid team" (Dudzinski & Frohoff 2008), possibly increasing the duo's chances with potential mates and decreasing confrontations with potential competition. Many studies show us that gre at distances may separate dolphins from each other in an open marine environment and water clarity is variable. In this kind of environment, vocal communication is most useful. Dolphins use whistles for group cohesion. Dolphins increase signature whistle production while separated (Smolker, Mann, & Smuts 1993; Watwood, Owen, Tyack, & Wells 2005). I ncreased production of signature whistles when dolphins are apart allows the dolphins to locate each other (Janik & Slater, 1998) Dolphins also use vocalizatio ns to advertise aggression. Studies suggest that more burst pulses are created during aggressive interactions than in other contexts (Overstrom 1983; Blomqvist & Amundin 2004) Not much research has been conducted on the social function of echolocation clicks, though it seems likely that they serve a purpose outside navigation and prey detection. While we have some idea how phonation production changes during different social si tuations, so far, no research has been conducted on phonation rates re lative to synchronous swimming. However, it is likely that vocalizations may be used to initiate and/or maintain synchrony between dolphins. The focus of the current study will address
Phonation rates and synchrony 26 t his issue by analyzing whistles, click trains, burst pulses and jaw claps in relation to synchronous swimming in a population of 4 captive males. Method Data collection occurred on November 14, 2008 with 4 male bottlenose dolphins residing at a captive location in central Florida The dolphins were Ranier (28 years), Khyber (17 years), Calvin (15 years), and Malabar (8 years). Simultaneous audio and video recordings were taken throughout the day. Wendi Fellner analyzed video recordings for synchronous be havior, and audio files were sent to Caitlin O Brien for analysis. Video was recorded to a MiniDV tape. Separate camcorders were used for each of three locations, and all clocks were synchronized manually to the computer Video from A and B pools (see Figure 1 below) was digitized in 5 minute segments that matched associated 5 minute audio files. The videos were digitized via Firewire (IEEE1394) and using Adobe Premier 6.0 with an NTSC DVtape format at 30 frames per second at the size of 720x480 pixels. Video files were saved in avi format. Figure 1. Tank set up: Hydrophone locations are identified with circles and video camera locations are identified with symbols of cam eras.
Phonation rates and synchrony 27 Three hydrophones (Cetacean Research Technology C54XRS, no filter with a flat frequency response up to 50kHz UltraSound Gate Gain = 4) were placed in the tank. One was located in the main tank to the side of the B pool gate. The second and third hy drophones were placed in each o f the back pools (A and B, see F igure 1). In addition, video cameras were placed to get above water views of both back pools while main pool activity was recorded on an underwater video camera. Data Coding Noldus Observer XT was used for video analyses. Wendi Fellner coded synchronous episodes in terms of subject, object, start sync, terminate sync, and out of view. The subject could refer to an individual dolphin or a group of dolphins, and referred to the dolphin(s) responsi ble for initiating a bout of synchrony or for swimming out of view. The object was the dolphin that was joined by the subject to begin a bout of synchrony. For example, if Calvin was swimming laps in A pool while Ranier was in the main tank and Ranier came into A pool and began swimming with Calvin, Ranier would be the subject and Calvin would be the object. Start sync was used to signify the beginning of a synchronous bout and was also used whenever a new dolphin joined an existing bout, while terminate sy nc was used whenever an individual dolphin or a pair broke synchrony. The code out of view was used either when a dolphin was not present in A or B pool at the start of the session or when a dolphin swam out of the back pools into the main tank. Before the study, the researchers were trained to mark the different phonations for type and length and a reliability check showe d good agreement (95 %). Audio files did not only cover synchronous episodes and could contain phonations the dolphins made in a varie ty of situations. File s lasted five minutes in length and contained some synchrony as
Phonation rates and synchrony 28 well as some asynchronous behavior. These asynchronous periods were used as a base rate comparison. Two additional files containing no synchrony were also used to assess a base rate. Avisoft SASLab Pro was used for acoustic analyses Phonations were reviewed via spectrogram and could be categorized as: whistles, echolocation click trains burst pulse sounds, jaw claps, or other. Figure 2, below, shows spectrograms of these phonation types. The coding scheme included information about the sound element type (BP = burst pulse, C = click, W = whistle, J = jaw clap, O = other) as well as the type of combination from which the sound wa s drawn. For example, a W element might be embedded in a long click train (WC). In addition, the channel in which the signal was strongest was noted, as well as any dolphins present in that location. This allowed for phonations to be attributed to smaller groups and sometimes to individual dolphins. Whistles were further subcategorized by frequency contour by Jenna Clark Burs t Pulse Burst Puls e with Jaw Clap Clicks Whistle Whistles within Clicks Figure 2. Phonation types. Horizontal axes are on identical time scales, and vertical axes measure phonation frequency to 30 kHz. Colors reflect sound intensity, with red signifying areas of greatest intensity.
Phonation rates and synchrony 29 Results Synchrony Synchrony was common. It occurred 26% of the time in files that contained some synchrony. There were a total of 28 syn chronous bouts between dyads that ranged from 0.8 3 seconds to 94 .27 seconds ( M = 25.06, SD = 25.94). The total number of bouts drops to 25 if dyads are combined in group swims. A group swimming event occurred when two males began swimming synchronously together and were then joined by a third dolphin. At no point during t he recording period did all 4 males swim together. In addition, Malabar was never part of a trio swim. Calvin participated in the majority of bouts (92%) while Malabar participated in fewest (28%). All dolphins except Malabar initiated more bouts than they terminated. Figure 3. Participation in Bouts of Synchrony 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% 90.00% 100.00% CALVIN KHYBER MALABAR RANIER % OF I NSTANCES IN R OLE D OLPHIN N AME % P ARTICIPATE % I NITIATE % T ERMINATE
Phonation rates and synchrony 30 Phonation Rates A total of 645 phonations were extracted and then categorized by type, potential producer(s), and whether they occurred during synchrony. Some vocalizations (8%) occurred when a subset of dolphins were swimming synchronously in the same pool with others, so it was not possible to determine who was phonating These phonations were removed from further analyses. Phonation rates were calculated to reflect the number of ph onation s per minute per dolphin. Minutes were calculated by measuring the number of seconds a dolphin or dolphins were present in A and B pools for each synchrony category: during synchrony (631.35 s), non synchronous periods in files with synchrony (1755.65 s), and non synchronous times from the two base rate files (495 s). These analyses revealed that phonations varied with synchrony. Synchronous pairs never produced burst pulse sounds or jaw claps during the swim, though they did use whistles and clicks. Figure 4. Phonations Per Dolphin Per Minute 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 BP C J O W P HONATIONS P ER D OLPHIN P ER M INUTE P HONATION T YPE S YNC D OLPHINS N ON -S YNC P ERIODS B ASE R ATE
Phonation rates and synchrony 31 Whistle Contours Signa ture whistles were rare overall Within 48 minutes of analyzed phonations, 287 whistle contours were extracted. Only 9.8% were signature whistles, and these did not reliably precede synchr ony. In only one instance in one synchronous bout with Calvin and Malabar, the "Calvin" signature was whistled by one of the pair members Otherwise, signature whistles were only produced outside of synchrony. For example, Ranier whistled his signature mul tiple times when Calvin and Khyber swam synchronously without him Khyber's signature was also only whistled outside of synchrony. In addition, two common contours, "hills" and "upcurve were never whistled during synchrony Figure 5. Whistle Contour Per Dolphin Per Minute 0.000 0.100 0.200 0.300 0.400 0.500 0.600 0.700 C ALVIN K HYBER R ANIER BUMP FLAT HILLS INTERSECTION R TRAPEZOID ULTRA BLIP ULTRA UP UPCURVE UPSWEEP # W HISTLES PER D OLPHIN PER M INUTE C ONTOUR N AME ( SIGNATURE WHISTLES AND OCCURANCES OF 10+) B ASE R ATE D URING N ON -S YNC O THER ( DURING SYNC ) S YNC D OLPHINS
Phonation rates and synchrony 32 Discussion In the files reviewed for the present study, phonations varied in synchronous vs. non synchronous contexts For example, n o burst pulse sounds or jaw claps occurred between synchronous animals Previous studies (Overst rom, 1983, Blomqvist & Amundin, 2004) suggest a correlation between burst pulses and jaw claps and aggressive behavior Absence of these types of phonations support the idea that synchrony serves an affiliative purpose. Synchronous dolphins also almost never produced signature whistles. Signature whistle research provides strong data for their use as cohesion calls. Therefore it was hypothesized that dolphins might coordinate synchrony via signature whistles. However, c ontrary to expectations, signature whistles did not typically precede synchrony. The production of Calvin's signature while he was engaged in synchrony raises the possibility that dolphins might maintain contact with other dolphins in the group while swimming with others. That Ranier repea tedly whistled his signature while Calvin swam synchronously with Khyber supports this claim In fact, s oon after the synchronous bout between Calvin and Khyber, Calvin initiated a bout of synchrony with Ranier. Therefore, it is possible that some bouts of synchrony are initiated via signature whistles. However, the current data suggest that this may be the exception rather than the rule. Inferential s tatistics could not be conducted with this sample, so variations in numbers of click trains and wh istles across contexts are hard to interpret However, these data do suggest that whistles and clicks are more common in non synchronous periods in which synchrony occurs vs. during times with no synchrony at all It is possible that periods with no synchrony at all are generally non socially interactive periods A s noted in Quick
Phonation rates and synchrony 33 and Janik (2008), wild dolphins whistle more frequently during social behavior than other types of behavior such as travelling. Xitco and Roitblat's findings (1996) suggest one possible reason for the decrease in clicks during synchrony rela tive to non synchronous periods. I f two dolphins swim close together, as they often do during synchrony, it would be possible for one to eavesdrop on its clicking partner making it unnecessary to clic k on its own. Two echolocators could also produce interference and uninterpretable echoes. This study utilized a small, captive, all male group of dolphins. T his group composition is unusual in captive facilities and further research should build on the current method to expand the data with this group In addition, c hanges in group composition may have yielded different results and offer an interesting follow up to this work Additional video cameras on the main tank would also strengthen the data reg arding synchrony because behavior in all tanks could be captured. While the presence of three hydrophones made it possible to narrow down the origin of phonations, it would have been beneficial to attribute vocalizations to individuals more frequently. Researchers have attempted to create such technology in the past, such as the vocalight (Tyack & Recchia 1991) but no device has yet been created that could reliably indicate a vocalizer for both narrow and broadband soun ds in very close proximity with others Further efforts should be made in this area as such a device would greatly enhance our understanding of the uses of dolphin phonations It is possible that this study is representative of the vocal behavior of only these four dolphins. T hese animals are housed in a relatively small area when compared with wild dolphins, and water clarity at this facility is excellent. These two factors raise the possibility
Phonation rates and synchrony 34 that vocal communi cation may not be as necessary for t he se animals as it would be for wild populations. Visual signals would be more readily available here than in murkier water. Indeed, research has shown that dolphins in clearer water produce fewer w histles than those in a cloudy environment (Dudzinski & Fr ohoff, 2008). Therefore, it would also be interesting to investigate the relation between visual and tactile signals and synchrony as well. Time spent swimming synchronously accounted for approximately a quarter of total time in files containing synchrony Synchrony may encompass a greater amount of time in wild populations where it might be more functional. For example, males at this facility do not have access to females and therefore do not need to engage in herding behavior like males in Shark Bay Whi le aggression between conspecifics does occur, risk of predation by sharks or orcas is nonexistent for these males. In addi tion, food is easily obtainable making coordinated foraging behavior unnecessary. The relative ease and safety of a c aptive lifestyl e raise interesting questions for synchrony and acoustic communication. While it is clear that these males are not benefitting from the se behavior s in the same ways wild dolphins might they still engage in synchrony with regularity. Therefore, synchrony may hold some other value such as promoting socials affiliation s or cooperation. This possibility is also supported by the differences in phonation production rates during synchrony and non synchrony. Burst pulses and jaw claps, phonations commonly associ ated with aggression, never occurred during synchrony. In addition, synchronous dolphins produced both whistles and clicks Production of both types of phonations increased during non synchronous periods. In a location with such clear underwater visibili ty, it seems unlikely that these dolphins rely to o
Phonation rates and synchrony 35 heavily on echolocation click trains to navigate the environment. Their increased production during non synchronous periods around synchrony hints at potential social uses of these phonations. Dolphins ha ve developed a complex communication system perfect for an u nderwater environment. When water clarity is poor an d visual signals become unavailable acoustic communication allows for clear signal transmission. However, a s we have seen dolphins continue to produce sounds with regularity even when other types of communication might suffice. How these sounds function continue s to be a mystery worth pursuing.
Phonation rates and synchrony 36 References Au, W.W.L., & Snyder, K.J. (1980 ) Long range target detection in open waters by an echolocating Atlantic bottlenose dolphin ( Tursiops truncatus ). Journal of the Acoustic Society of America, 68 (4), 1077 1084. Blomqvist, C. & Amundin, M. (2004). High frequency burst pulse sounds in agonistic/aggressive interactions in bottlenose dolphins, Tursiops truncatus In J.A. Thomas, C.F. Moss, & M. Vater (Eds.), Echolocation in bats and dolphins ( pp. 425 431 ) Chicago, IL: The University of Chicago Press. Connor, R.C., Smolker, R.A., & Richards, A.F. (1992). Two levels of alliance formation among male bottlenose dolphins ( Tursiops sp. ). Proceedings of the National Academy of Sciences of the United States of America, 89, 987 990. Connor, R. C., & Smolker, R.A. (1996). Pop' goes the do lphin: a vocalization male bottlenose dolphins produce during consortships. Behaviour, 133, 642 662. Connor, R.C.; Heithaus, R.M. and Barre, L.M. (1999). Super alliance of bottlenose dolphins. Nature 371 : 571 572 Connor, R.C., R. Wells, J. Mann, and A. Read. ( 2000). The bottlenose dolphin: S ocial relationships in a fission fusion society. In J. Mann; R.Connor, P.Tyack, and H. Whitehead (Eds.), Cetacean Societies: Field studies of whales and dolphins (pp 91 126). Chicago, IL: University of Chicago Press. Connor, R.C., Heithaus, M.R., & Barre, L.M. (2001). Complex social structure, alliance stability and mating access in a bottlenose dolphin super alliance'. Proceedings: Biological Sciences, 268, 263 267. Connor, R.C., Mann, J., & Watson Capp, J. (2006). A sex specific affiliative contact behavior in Indian Ocean bottlenose dolphins, Tursiops sp. Eth ology 112 631 638. Dudzinski, K.M., & Frohoff, T. (2008). Dolphin Mysteries: Unlocking the Secrets of C ommunication New Haven: Yale University Press. Duffield D.A., & Wells R.S. ( 1991 ) The combined application of chromosomal protein and molecular data for the investigation of social unit structure and dynamics in Tursiops truncatus Representatives of the International Whale Commission (special issue), 13 155 170. Fellner, W., Stamper, S.A., Losch, B.A., Dahood, A., & Bauer, G.B. Abstract. The development of synchrony and behavior in the bottlenose dolphins. 16 th B iennial Conference of the Biolog y of Marine Mammals, December 2005, San Diego, CA.
Phonation rates and synchrony 37 Fertl, D., & Leatherwood, F. (1997). Cetacean interactions with trawls: a preliminary review. Journal of Northwestern Atlantic Fisheries Science 22 219 248. Gazda, S., R.C. Connor, R.Edgar, & F. Cox. ( 2005 ) A Division of Labor with Role Specialization in Group Hunting Bottlenose Dolphins ( Tursiops truncatus ) off Cedar Key, Florida. The Proceedings of the Royal Society: Biological Sciences, 272 135 140. Hinde, R.A., & Atkinson, S. ( 1970 ) Assessing the roles of social partners in maintaining mutual proximity, as exemplified by mother infant relations in rhesus monkeys. Animal Behaviour, 18 169 176. Hoese, H.D. (1971). Dolphin feeding out of water in a salt marsh. Journal of Mammology, 52, 222 223. Janik, V. M. (2000). Food related bray calls in the wild bottlenose dolphins ( Tursiops truncatus ). Proceedings: Biological Sciences, 267, 923 927. Janik, V.M. (2000). Whistle matching in wild bottlenose dolphins ( Tursiops truncatus). Scie nce, 289 1355 1357. Janik V.M., & Slater P.J.B. (1998) Context specific use suggests that bottlenose dolphin signature whistles are cohesion calls. Animal Behaviour, 56 829 838. Kuczaj, S.A., Makecha, R., Trone, M., Paulos, R.D., & Ramos, J.A. (2006). Role of peers in cultural innovation and cultural transmission: Evidence from the play of dolphin calves. International Journal of Comparative Psychology, 19, 223 240. Madsen, P.T., Kerr, I., & Payne R. (2004) Echolocation clicks of two free ranging, oceanic delphinids with different food preferences: false killer whales Pseudorca crassidens and Risso's dolphins Grampus griseus. The Journal of Experimental Biology, 207, 1811 1823. Mann, J., & Smuts, B. (1999). Behavioral developm ent in wild bottlenose dolphin newborns ( Tursiops sp. ). Behaviour, 136, 529 566. Noren, S.R., Biedenbach, G., Redfern, J.V., & Edwards, E.F. (2008). Hitching a ride: the formation locomotion strategy of dolphin calves. Functional Ecology, 22, 278 283. Overstrom, N.A. (1983). Association between burst pulse sounds and aggressive behavior in captive Atlantic bottlenosed dolphins ( Tursiops truncatus ). Zoo Biology, 2, 93 103. Pack, A.A., Herman, L.M., Hoffman Kuhnt, M., & Branstetter, B.K. (2002). The o bject
Phonation rates and synchrony 38 behind the echo: dolphins ( Tursiops truncatus ) perceive object shape globally through echolocation. Behavioural Processes, 58, 1 26. Parsons, K.M., Durban, J.W., Claridge, D.E., Balcomb, K.C., Noble, L.R., & Thompson, P.M. (2003). Kinship as a basis for alliance formation between male bottlenose dolphins, Tursiops truncatus, in the Bahamas. Animal Behaviour, 66, 185 194. Perelberg, A. & Schuster, R. (2008). Coordinated breathing in the bottlenose dolphins ( Tursiops t runcatus ) as cooperation: integrating proximate and ultimate explanations. Journal of Comparative Psychology, 122, 109 120. Pressman, E (2007). Acoustic production rates in a group of captive bottlenose dolphins (Tursiops truncatus) over population cha nges. Unpublished undergraduate thesis, New College of Florida, Sarasota, FL, USA. Quick, N.J., & Janik, V.M. (2008). Whistle rates of wild bottlenose dolphins ( Tursiops truncatus ): Influences of group size and behavior. Journal of Comparative Psychology 122 305 311. Quintana Rizzo, E., Mann, D.A., & Wells, R.A. (2006). Estimated communication range of social sounds used by bottlenose dolphins ( Tursiops truncatus ). Acoustical Society of America, 1671 1683. Reynolds, J.E., Wells, R.E., & Eide, S.D. (2000). The Bottlenose D olphin : Biology and Conservation Gainesville: University Press of Florida. Rossbach, K.A., & Herzing, D.L. (1997). Underwater observations of benthic feeding bottlenose dolphins ( Tursiops truncatus ) near Grand Bahama Island, Bahamas. Marine Mammal Science, 13, 498 504. Samuels, A., & Gifford, T. (1997). A quantitative assessment of dominance relations among bottlenose dolphins. Marine Mammal Science, 13, 70 99. Silber, G.K., & Fertl, D. (1995). Intentional beaching by bottlenose dolphins ( Tursips truncatus ) in the Colorado River delta, Mexico. Aquatic Mammals, 21, 183 186. Smolker, R.A., Mann J., & Smuts B.B. (1993) Use of signature whistles during separations and reunions by wild bottlenose dolphin mothers and infants. Behavioral Ecology and Sociobiology, 33 (6), 393 402. Tamaki, N., Morisaka, T., & Taki, M. (2006). Does body contact contribute towards repairing relationships? The association betw een flipper rubbing and aggressive behavior in captive bottlenose dolphins. Behavioural Processes, 73, 209 215. Tyack, P.L., & Recchia, C.A. (1991). A datalogger to identify vocalizing dolphins. Journal of the Acoustic Society of America, 90 (3), 1668 16 71.
Phonation rates and synchrony 39 Watwood, S.L., Tyack, P.L., & Wells, R.S. (2004). Whistle sharing in paired male bottlenose dolphins, Tursiops truncatus. Behavioral Ecological Sociobiology, 55, 531 543. Watwood, S.L., Owen, E.C.G., Tyack, P.L., & Wells R.S. (2005) Signature whistle use by temporarily restrained and free swimming bottlenose dolphins, Tursiops truncatus. An imal Behaviour, 69 (6), 1373 1386. Wells, R. S., M. D. Scott & A. B. Irvine. ( 1987 ) The social structure of free ranging bottlenose dolphins. Current Mammalog y, 1, 247 305. Xitco M.J., & Roitblat H.L. (1996) Object recognition through eavesdropping: Passive echolocation in bottlenose dolphins. Animal Learning & Behavior, 24 (4), 355 365.