Vocal repertoire of South American fur seals, Arctocephalus australis : structure, function, and context

Vocal repertoire of South American fur seals,Arctocephalus australis: structure, function, andcontext

Alana V. Phillips and Ian Stirling

Abstract: We describe the vocal repertoire of male and female South American fur seals (Arctocephalus australis)breeding at Punta San Juan, Peru, the first such description for any member of the genus. We distinguished 11 calltypes, which we grouped into four functional classes: investigative, threat, submissive, and affiliative calls. Barking isused during non-agonistic investigation of other individuals. Threat calls of South American fur seals show gradation ofstructure, form, and apparent meaning, and are grouped into two series: nontonal or respiratory sounds, and pulsed orguttural sounds that sometimes include a terminal tonal component. This might be adaptive in enabling display behav-ior to be flexible in agonistic situations, allowing participants to interpret situations on the basis of contextual cues andtheir own physical ability and experience. In contrast, vocal displays such as submissive, full-threat, and affiliative callstend to have a discrete acoustic structure. Of these, full-threat, female-attraction, and pup-attraction calls share acousticcharacteristics: all are long, loud calls composed of both pulsed and tonal components, and show sufficient variation toallow individual recognition. We attempt to establish a base line for standardizing nomenclature and acoustic analysis,to facilitate further comparative research on the vocal repertoire ofArctocephalusspecies.

Résumé: On trouvera ici la description du répertoire vocal des mâles et des femelles de l’Otarie à fourrure australe(Arctocephalus australis) au cours de leur période de reproduction à Punta San Juan, Pérou; il s’agit là d’une descrip-tion inédite chez une espèce de ce genre. Nous avons reconnu 11 types de cris que nous avons regroupés en quatreclasses fonctionnelles : cris d’investigation, cris de menace, cris de soumission et cris d’affiliation. Les jappements sontémis lors de l’investigation non-agonistique d’autres individus. On reconnaît une gradation de structure, de forme et designification apparente dans les cris de menace de l’Otarie à fourrure australe qui peuvent être regroupés en deux séries :sons atonaux ou respiratoires et sons en impulsions ou gutturaux qui se terminent parfois par une composante tonale.Peut-être faut-il voir là une caractéristique évolutive qui permet des comportements de parade flexibles dans les situationsagonistiques où les participants peuvent interpréter les situations d’après le contexte, en fonction de leurs propres capacités etexpériences physiques. En revanche, les manifestations vocales comme les cris de soumission, de menace directe etd’affiliation ont une structure acoustique discrète. Parmi ces cris, les cris de menace directe, les cris d’appel aux femelleset les cris d’appel aux petits ont des caractéristiques acoustiques communes : tous ces cris sont forts et longs; ils comportentdes composantes à impulsions et des composantes tonales et ils sont suffisamment distinctifs pour permettre la reconnais-sance individuelle. Nous tentons d’établir un système de base pour l’uniformisation de la nomenclature et l’analyseacoustique, de façon à faciliter la recherche comparative sur le répertoire vocal des espèces d’Arctocephalus.

[Traduit par la Rédaction] 437

Phillips and StirlingIntroduction

The genusArctocephalus(Pinnipedia: Otariidae) includeseight extant species of southern fur seal that inhabit a rangeof climates from subantarctic to tropical. Currently, thesespecies are distinguished largely on the basis of discontinuitiesin their geographic ranges, and secondarily on morphometricmeasurements and behavioral characteristics such as airbornevocalizations. Even so, the interrelationships of the eightspecies are still under review (Repenning et al. 1971; Stirlingand Warneke 1971; Berta and Deméré 1986; Lento et al.1997; Wynen et al. 2000). The current geographic ranges of

some species overlap, and hybridization has been observedbetween the subantarctic fur seal (A. tropicalis), New Zea-land fur seal (A. forsteri), and Antarctic fur seal (A. gazella)on Macquarie Island (Goldsworthy et al. 1999), somethingthat may not have occurred prior to their economic extinc-tion in the 1800s and subsequent recovery from remnantpopulations.

Vocalizations associated with territorial behavior, repro-duction, or mother–young recognition may act as isolatingmechanisms that keep species discrete, and comparisons ofstudies of various species of southern fur seals suggesthomologies in call types (Stirling and Warneke 1971; Roux

Can. J. Zool.79: 420–437 (2001) © 2001 NRC Canada

420

DOI: 10.1139/cjz-79-3-420

Received January 31, 2000. Accepted November 17, 2000. Published on the NRC Research Press Web site on March 5, 2001.

A.V. Phillips.1,2 Department of Biological Sciences, University of Alberta, Edmonton, AB T6G 2E9, Canada.I. Stirling. Canadian Wildlife Service, 5320 122 Street, Edmonton, AB T6H 3S5, Canada, and Department of Biological Sciences,University of Alberta, Edmonton, AB T6G 2E9, Canada.

1Corresponding author (e-mail: [email protected]).2Present address: The Marine Mammal Center, Marin Headlands, 1065 Fort Cronkhite, Sausalito, CA 94965, U.S.A.

J:\cjz\cjz79\cjz-03\Z00-219.vpWednesday, February 28, 2001 1:34:00 PM

Color profile: DisabledComposite Default screen

and Jouventin 1987; Miller 1991; Page et al.3). Because furseals produce a range of distinctive and function-specific vo-calizations, it is possible to undertake quantitative compara-tive studies of the structure and function of the same callemitted by different species, even if the mechanisms bywhich ecological factors may have influenced their evolu-tion are unknown. Furthermore, Miller (1991) suggested thatcertain pinniped vocalizations may exhibit structural grad-ing; that is, calls show “extensive continuous variation andintergradation of categories.” If the vocalizations emitted byfur seals are graded, the behaviors associated with the callsmight exhibit a parallel pattern of gradation in context andfunction (Morton 1982). Thus, the use of vocalizations andassociated behavior as a way of comparing the behavior, andpossibly the evolution, of the southern fur seals, requiresthat the vocal repertoire of all species, as well as the func-tions of the calls and the contexts in which they are given,be quantitatively described in a comparable manner.

In comparing the vocalizations of the southern fur seals, itis particularly important to document the vocal behavior ofthe South American species (A. australis) because it is thoughtto be the least derived member of the genus (Repenning etal. 1971) and therefore its vocal repertoire is likely more re-flective of the ancestral condition than that of any of theother species. Moreover, South American fur seals breedingin Peru face a number of challenges, including crowding to adegree that might not have been the original condition forthis species, owing to long-term disturbance and huntingpressure by humans that limited the protected habitat suitablefor breeding. In addition, pups of this species are subject topredation by South American sea lions (Otaria byronia), andhigh daytime temperatures that cause females to make dailymovements through the colony to the sea for cooling. Thecombination of these factors has resulted in levels of aggres-sion among females that are the highest reported for anyotariid species (Harcourt 1991, 1992; Majluf 1987, 1992).Thus, the vocal repertoire of this species might be expectedto include a range of threat vocalizations, and because agonisticencounters occur frequently, calls are expected to show somegradation of form and function (Miller 1991).

To date, descriptions of the vocal behavior of the SouthAmerican fur seal have been anecdotal (Vaz-Ferreira 1956,1971; Trillmich and Majluf 1981). In this paper, we quanti-tatively document the vocal repertoire of the South Ameri-can fur seal in Peru during the breeding season, describe thefunctions of the calls and the contexts in which they aregiven, and compare them with those reported for all otherArctocephalusspecies. We present the results of this studyas the base line for the vocal repertoire forA. australisaswell as a methodological template for future comparative re-search on other members of the genus.

Materials and methods

Punta San Juan (15°22′S, 75°12′W) is one of the main breedingsites for South American fur seals in Peru and has been describedelsewhere (Trillmich and Majluf 1981; Majluf 1987). We con-ducted our study during the breeding seasons of November–December 1994 and October–December 1995, corresponding to

the annual peak of pupping and mating activity (Majluf 1987,1992). Behavioral observations and acoustic recordings were madeat beach N4, from a clifftop observation point downwind of thecolony. The study colony numbered 150–200 adult fur seals at thepeak of the breeding season.

Categorization and terminology of callsPreliminary observations of vocalizations were conducted ad li-

bitum from a clifftop bench at beach N4 and during tagging activi-ties on beaches N4 and S3. Call types were classified by ear in thefield, making maximum use of previous terms from the literatureand our experience with otherArctocephalusspecies (e.g., Stirling1971a; Stirling and Warneke 1971; Miller 1974). With the exceptionof the pup-attraction call (see below), we differed from Stirling andWarneke (1971) and Trillmich and Majluf (1981) by not categoriz-ing calls as “male” or “female.” Although we attempted to assigncall names according to structural characteristics rather than per-ceived function, we retained some function-based names that werealready prevalent in the literature (for example, pup-attraction calls,full-threat calls, submissive calls) (Phillips 1998).

Miller (1991), Miller and Job (1992), and Insley (1992) have ar-gued that “pup-attraction call” and “female-attraction call” may bemisrepresentative, particularly when pairs call during physical con-tact, and may exclude interactions between mothers and yearlings.“Female-attraction call” is especially ambiguous, since it distin-guishes neither the sender (which could be an adult male attempt-ing to copulate) nor the recipient (which could be any female ofany age). In this paper, we use “pup-attraction call” (PAC) to referto non-agonistic calls made by mothers seeking or trying to attracttheir offspring (whether pup, yearling or 2-year-old). We use“female-attraction call” (FAC) to refer to non-agonistic calls madeby pups or juveniles towards females that they appear to be tryingto identify as their mother.

Acoustical analysesAfter the call types had been identified, we used spectrographic

analysis to describe the calls quantitatively. Tape recordings weremade using a Marantz PMD430 cassette recorder and a SennheiserK3N/ME88 directional microphone with the filter set to PositionIII to reduce wave and wind noise; this system had a combined fre-quency response of 50 Hz – 18 kHz. Most of the calls were re-corded from the clifftop observation point, approximately 15 mfrom the animals, allowing us to make good-quality recordingswithout alerting the animals to our presence. Additional vocaliza-tions were recorded directly during tagging activities on the beach,from distances ranging from 0.1 to 3.0 m. These included some ofthe lower frequency threat calls that were difficult to record from adistance and many of the calls made between mothers and pups asthey maintained contact during these temporary disturbances. Sub-sequent acoustical analyses showed that our presence did not affectthe structural characteristics of these calls (Phillips 1998). After re-viewing the tapes, we selected up to 20 of the best quality calls ineach category (using calls from as many different individuals aspossible), or up to 20 calls per individual for full-threat calls (FTCs),PACs, and FACs (Phillips and Stirling 2000).

Sound spectrograms were prepared using SIGNAL/RTS (EngineeringDesign, Belmont, Mass.). Calls were sampled at a rate of 20 kHzover the frequency range 0–10 kHz. Spectrograms were calculatedfrom 512-point fast Fourier transforms, with a corresponding fre-quency bandwidth of 39 Hz. Some calls contained wideband extra-neous noise that extended into frequency ranges higher than thecutoff at 12 kHz; however, noise above this level was not consid-ered a significant contribution to the overall sound of the call

© 2001 NRC Canada

Phillips and Stirling 421

3B. Page, S.D. Goldsworthy, and M.A. Hindell. Inter-specific differences in male vocalizations of three sympatric fur seals (Arctocephalusspp.). Submitted for publication.



© 2001 NRC Canada

422 Can. J. Zool. Vol. 79, 2001

because of the effect of masking by background noise from thecolony (Moore and Schusterman 1987; Renouf 1991).

For each call we measured the following acoustic characteristicsusing methods described in detail in Phillips and Stirling (2000):duration (ms), number of parts per call, harmonic interval (Hz),frequency of the lowest visible harmonic at the onset, maximum,and end of the call (Hz), period (ms) and range (Hz) of rhythmicfrequency modulation (FM), when present, and the frequency ofthe first energy peak (Hz) of the call. In addition, we measured theintercall interval (ms) for calls that occurred in bouts (barking andpuffing), and the interpulse interval (ms) for calls in which pulsescould be distinctly heard (growls, low-intensity threat calls, andguttural threat calls (GTCs)). Finally, we also measured the maximumamplitude (V) of barks and threat calls. In bioacoustics, amplitudemeasurements can be affected by numerous factors, including theanimal’s emotional state, its distance from and orientation to themicrophone, background-noise level, and battery charge of the re-cording equipment (Phillips and Stirling 2000). Therefore, we esti-mated relative amplitude by measuring its maximum value on theamplitude waveform for each call. We treated these values as rela-tive within this study; caution should be used when making com-parisons with other studies because of differences in recordingsituation and equipment used.

Behavioral observationsWe made behavioral observations to verify the context of each

call and the situation in which it was produced, and to eliminatethe bias associated with assigning function-based names to certaincalls. We counted all calls, and categorized their context, for alladult males and females within a specific focal area of approxi-mately 135 m2 (a “focal subgroup” sensu Altmann 1974). It waspossible to continuously monitor all individuals in the samplingarea throughout each sampling period. This way, counts of callswere independent of the sex, age-class, and behavior of the emit-ting animals, and we were not biased towards the more conspicu-ous individuals or behaviors, nor were we apt to overlook calls thatwere subtle or had no apparent purpose (Altmann 1974). Samplingwas conducted for 1-h periods within daily time windows of06:30–08:30, 11:30–13:30, and 16:00–18:00. We collected a totalof 220 h of observations at beach N4 (72 h in 1994 (November12 – December 12) and 148 h in 1995 (September 27 – December11)). Adult male and female fur seals were counted at the begin-ning of each sampling period to enable a standardized measure ofcalling rate (number of calls per individual per hour) to be calcu-lated. This number was a relative measure, since it was not possi-ble to monitor the amount of migration into or out of the samplingarea during the observation period. Over the 220 h of sampling, thenumber of males within the sampling area at the start of the obser-vations ranged from 1 to 8 (median = 3); that of females rangedfrom 8 to 49 (median = 21). Barks and puffs were produced insuch rapid succession that it was not possible to count individualcalls. For the purpose of this analysis, therefore, each series ofthese vocalizations separated from the next by a distinct period ofsilence was counted as one “call unit.”

Each call or call unit counted was classified first by the sex ofthe emitting animal, second by structural type, and third by thetype of individual to whom it was made. Calls that were not obvi-ously directed at any particular individual, or that appeared to bedirected towards a group of animals, were defined as calls made“to the colony.” Calls that could not be classified by structure or re-cipient were recorded as “unclassified.” When vocalizations weremade to other species, the type of call and recipient species werenoted, but because these calls were produced so infrequently, wepooled them as “unclassified” for subsequent analyses.

Statistical analysesObservations in the field indicated that mother fur seals’ behav-

ior changed after there were at least 25 pups present. For example,above this threshold, mothers were much less likely to abandon theirpups when humans approached the colony (Majluf and Goebel1992). Thus, the sampling periods were divided a posteriori intothree intervals, defined by the level of pupping activity on thebeach:no pups, onset (1–25 pups present), andpeak (>25 pupspresent) (Phillips 1998). In 1994, sampling was carried out onlyduring the peak period. In 1995, data were collected during allthree periods (n = 52, 33, and 63 h, respectively). Since the pres-ence and number of pups necessarily had an effect on female call-ing, data from each period in 1995 were analyzed separately, andonly those from the peak period in 1995 were compared with the1994 data.

We conducted a series of likelihood-ratio tests (Monte Carlo,5000 replications) of the hypothesis that the way in which adult furseals use each call type depends on the sex–age class of the recipi-ent. For adult males and females, and for each period of the breed-ing season, we analyzed a matrix of 10 call types and 6 recipientclasses (both variables included an additional category of “unclas-sified” call types or recipients). Randomization techniques weremore appropriate to use than conventional methods because callsdid not occur in every cell of each matrix, and because these ran-domization tests minimize the effect of pooling samples (Crowley1992; Adams and Anthony 1996). Statistical analyses were madeusing StatXact (Cytel Software Corp., Cambridge, Mass.).

Results

At Punta San Juan, both male and female South Americanfur seals produce a large range of vocalizations. We distin-guished 11 call types, which may be grouped into fourfunctional classes: investigative, threat, submissive, andaffiliative. These are summarized in Table 1, together withalternative terms used in other studies ofArctocephalusspp.

In all periods of the breeding season, the type of call usedby adult South American fur seals depended on the recipientof the call (Table 2). Both frequency and distribution of callswere similar during peak periods in both seasons (Tables 3,4). Since the peak period was most relevant to the rest of ouranalyses, we present only those data when referring to thefrequency and distribution of calls in the following sections,except where noted (data for the periods prior to and duringthe onset of pupping are summarized in Phillips 1998). Callsmade by young fur seals were not quantified, so the informa-tion presented here on call usage by pups and juveniles issummarized from field notes only.

Calls made to other species were rare, accounting for 0.55and 0.15% of adult male and female calling, respectively(Table 5). For adult males, interspecific calling appeared tobe related to defence against South American sea lions, whilemother fur seals most often threatened approaching turkeyvultures (Cathartes aura).

Investigative calls

BarkingThis is a series of low-energy, rapid exhalations of air

through the nose or slightly open mouth. On a spectrogram,barking appeared as a series of pulses of wideband noiseproduced at a rate of approximately 3–5/s. Although individ-ual barks were short (around 50 ms; Table 6), they wereproduced in series ranging from 1 s to several minutes in du-ration (Fig. 1a). No tonal components were present.



© 2001 NRC Canada


Vocalizationa Alternative name(s) Reference(s)

Investigation callBark Whimpering Bonner 1958, 1968; Miller 1971, 1974, 1975; Shaughnessy

et al. 1988Speaking Paulian 1964Honking, grunting, clucking Rand 1967Whickering Peterson et al. 1968; Roux and Jouventin 1987Whicker–bark Pierson 1987Aboiement Caudron 1991; Tollu 1982Resonancia nasal Vaz-Ferreira 1956

Threat callSnort Oral snorting Miller 1974

Renâcler Caudron 1991Puff Coughing Feldman 1993; Paulian 1964; Peterson et al. 1968; Rand 1967

Hissing Rand 1967Menace toussée Caudron 1991

Chuff Boundary puff Pierson 1987Growl Low-intensity threat Miller 1974; Pierson 1987

Grondement / grognement sonore Caudron 1991Gruñido Vaz-Ferreira 1971

Low-intensity threat call (LITC) Low roar Paulian 1964Short roar Stirling and Warneke 1971Low-threat call Trillmich and Majluf 1981Menace de faible intensité Tollu 1982Rugido bajo Vaz-Ferreira 1956Rugido grave Vaz-Ferreira 1971

Guttural threat call (GTC) Choke Stirling 1970; Paulian 1964(?)Guttural challenge Stirling 1971a; Stirling and Warneke 1971High-intensity guttural threat Stirling and Warneke 1971Provocation gutturale Tollu 1982Vocalisation gutturale saccadée Caudron 1991

Full-threat call (FTC) Roar Feldman 1993; Paulian 1964; Peterson et al. 1968; Rand 1967Trumpeted roar Miller 1971, 1974Territorial call Roux and Jouventin 1987Menace de forte intensité Tollu 1982Rugissement Caudron 1991

Appeasement callSubmissive call Howl Paulian 1964

Whine Stirling 1970Cri de soumission Tollu 1982Alarido Vaz-Ferreira 1956, 1971

Affiliative callPup-attraction call (PAC) Bleat Paulian 1964

Bellow or lowing Rand 1967Bawl Peterson et al. 1968Pup-contact call Trillmich 1981Cri d’appel du jeune Tollu 1982Cri de la mère Caudron 1991Balido Vaz-Ferreira 1971

Female-attraction call (FAC) Bleating Bonner 1968; Paulian 1964; Rand 1967Tantrum call Majluf 1987Cri d’appel de la mère Tollu 1982Cri du petit Caudron 1991Balido Vaz-Ferreira 1971

aCalls are arranged in presumptive functional classes.

Table 1. Call names used in this study and alternative names used in the literature forArctocephalusspp.



Context of barking—Barking was typically used during non-agonistic investigation of other individuals. The vibrissaewere erect and oriented forwards, and the emitting animalusually inclined or nodded its head towards the recipient(Fig. 2a). Barking to females represented 40–50% of allcalls made by males (Table 3). Males typically responded toagonistic interactions among females by approaching the fe-males and barking continuously. Males also barked duringcopulation and in response to other mild disturbances orstimuli such as mothers calling to their pups.

Adult females appeared to use barking in a similar way tomales, although rarely and only during interactions with otherfemales and juveniles (Table 4). Females sometimes barkedwhen an apparently nonthreatening individual approachedthem too closely.

Threat callsThese were grouped into two series of graded vocalizations:

nontonal or “respiratory” sounds (Peters and Wozencraft 1989)produced by forceful exhalation, and guttural or distinctivelypulsed sounds produced by vibration of the vocal cords.

Respiratory threat calls

Snort—This is a single sharp exhalation of air, given withthe force of a puff and intermediate in duration betweenpuffs and chuffs (Table 6). This sound consisted of high-intensity wideband noise; no pulsation or tonal componentswere present (Fig. 1b).

Puffing—This is a series of strong exhalations and inhala-tions produced in rapid succession at rates ranging from 4 to50/s (Table 6). Like barks, individual puffs were short (typi-cally less than 100 ms; Table 6) and were produced in serieslasting up to 2 min. Puffing was distinguished from barkingby medium to high levels of energy and a tendency towardspulsation (Fig. 1c).

Chuff—This is a high-intensity sound around 150 ms in du-ration (Table 6), produced by a loud, forceful exhalationfrom the larynx, often made as the animal threw itself downon its ventral surface. Chuffs were usually produced singlyor in very short series and were often immediately followedby sharp inhalations that appeared on a spectrogram as alower intensity pulse (Fig. 1d).

Guttural sounds

Growl—This is a low-energy, rapidly pulsed call with domi-nant frequencies around 400–700 Hz (Table 6, Fig. 1e).Growls typically consisted of one or two parts and variedconsiderably in duration, ranging from less than 1 s tomorethan 10 s (Table 6).

Low-intensity threat call (LITC)—This is a pulsed call ofmoderate duration (1–2 s) and moderate to high energy. Itwas distinguished from growls by its intensity and the tendencyfor wideband noise to overlay the dominant frequencies, whichwere somewhat higher than in growls. In addition, pulsestended to be produced less rapidly than in growls (Table 6,Fig. 1f ).

Guttural threat call—The GTC is a distinctive, slowly pulsed(around 55 ms between pulses), high-energy call with nonoisy components or tonal structure. Each pulse within thecall could be distinguished from the next on spectrograms(Fig. 1g) and by the human ear, to which the calls sounded(onomatopoetically) like a series of “guug-guug-guug” sounds.As a result, GTCs tended to exhibit greater call complexitythan growls or LITCs, averaging more than 10 parts per call(Table 6).

Full-threat call—The FTC is a high-energy call typicallycomposed of two regions: a distinctly pulsed, low-frequencycomponent with energy around 700 Hz, resembling a LITCor GTC, followed by a higher frequency tonal componentwith an average fundamental frequency of 1150 Hz (Table 6).Several harmonics were usually present, and the durationranged from 1 to 2 s. FTCs varied in structure between individ-uals andamong both adult males and females (Figs. 1i–1l;A.V. Phillips and I. Stirling, unpublished data and analyses).Because of the distinct pulses at the start of these calls, therewere often up to 18 parts per call (Table 6).

Context of threat callsThreat calls comprised 30–50% of calls made by both

males and females, the recipient most often being a memberof the same sex (Tables 3, 4). Males used the full range ofthreat calls when interacting with other males, but mainlythreatened females with growls and paid little or no attentionto pups and juveniles (Table 3). Adult females used growlsand puffs most frequently, rarely producing most of the higherintensity threat calls (Table 4).

Snorts were not included in the analysis of call usage be-cause they were produced infrequently and were difficult todetect. We observed snorts being made by territorial malesas an unfamiliar male approached and, on occasion, by chal-lenging males as they approached an occupied territory. Ofthe 34 snorts recorded during the 148 h dedicated to callquantification in 1995, 2 were directed to a juvenile sea lionthat was approaching the colony, 1 to a juvenile fur seal, andthe remainder to adult male fur seals. Neither adult femalesnor juveniles were observed to produce snorts.

Calls indicating a low to moderate level of aggression,such as growls and puffs, were used most frequently by bothadults (Tables 3, 4) and juveniles, although usage was notquantified for the latter. Adult males also produced LITCs(12–20% of threat calls); however, these were not observedfrom either adult females or juveniles. Growls, LITCs, and

© 2001 NRC Canada

424 Can. J. Zool. Vol. 79, 2001

Adult males Adult females

Period of season G2 df G2 df

1994Peak 1222 36 19850 35

1995No pups 785 32 2253 28Onset 755 35 4136 30Peak 1791 40 21860 28

Note: All tests are significant atp < 0.001.

Table 2. Likelihood-ratio tests of independence betweenthe type of call used by the sender and the recipient of thecall, for adult South American fur seals during each periodof the pupping season in 1994 and 1995.



puffs appeared to form a gradient of aggressive intent. Growlswere very low intensity threat calls that were not necessarilydirectly focused towards any one individual, nor associatedwith a particular posture. They appeared to be used in responseto the slightest perceived threat, including nearby turkey vul-tures and penguins passing through the colony (Table 5).LITCs were typically made with the head and body orienteddirectly towards the recipient, most often an intruding male,apparently conveying the potential for the level of threat toescalate. Puffing seemed to indicate a moderate level of inten-sity, and was often accompanied by oblique staring, open-mouth gaping, head shaking, and lunging (Fig. 2b). This wasthe first level at which physical contact such as biting wasseen to follow. Of female threat calls to other females, almost20% consisted of puffing, whereas females rarely puffed tomales (Table 4).

For adults of both sexes, growls, puffs, or LITCs alsoseemed to be effective in intimidating pups or juveniles sothat they vacated a space which they intended to use; indeed,this was the only context in which we observed adult malesvocalizing to young fur seals. Growls and puffs were oftenproduced by pups, yearlings, and juveniles (2- and 3-year-olds). Most of these occurred during play, such as mock ter-ritorial disputes or mock fighting. We also saw pups puffingin response to being approached by turkey vultures or hu-mans. Yearlings also growled in response to threatening adultmales and females.

High-intensity agonistic calls such as GTCs, chuffs, andFTCs were typically used during situations of acute aggres-sion, in which the animal seemed prepared to fight (Fig. 2c).Thus, the calls were infrequent and constituted 10–15% ofmale threats and less than 1% of female threat calls (Ta-bles 3, 4). Guttural threat calls were highly agonistic andwere not given as casually as growls or LITCs. While mak-ing this call, the head was typically oriented directly towardsthe threatening individual, with the neck outstretched in thealert posture (Stirling 1971a). Chuffs occurred rarely, typi-cally in boundary disputes between neighboring territorialmales. FTCs seemed to indicate the highest level of arousal,and males produced them twice as often as GTCs and chuffs(Table 3). Full-threat calls were made in response to a signif-icant perceived threat, such as during territorial conflicts be-tween males or by both sexes towards humans or adult sealions (Table 5) that approached the colony. FTCs were pro-duced either directly towards the recipient, like GTCs, orwith the neck stretched and nose pointed upwards while star-ing obliquely at the opponent. Both of these postures accen-tuate the size of the emitting animal and probably enhancethe effect of the display.

Males also made infrequent FTCs “to the colony,” possi-bly advertising their status to nearby males or females (Ta-ble 3). We typically observed this behavior in adult malesthat had just won a territorial dispute and ousted the previ-ous male. One male produced over 80 FTCs during the first

© 2001 NRC Canada


Recipient of call

Call type Female Male Pup Juvenile Colony Unclassified N

1994 (n = 72)Bark 39.0 3.6 — 0.2 1.4 5.1 884Threat calls (total) 8.4 34.6 — — 0.5 4.7 864

Puff 2.0 6.6 — — — 0.2 157Chuff — 1.7 — — — 0.1 32Growl 6.1 11.8 — — 0.1 1.3 346LITC 0.2 9.0 — — 0.3 2.0 206GTC 0.1 1.9 — — — 0.5 44FTC — 3.6 — — 0.1 0.7 79

Submissive calls — 0.1 — — — — 2Affiliative calls — 0.1 — — 0.3 0.1 9Unclassified 0.3 0.2 — — 0.4 1.8 37N 857 692 0 3 46 198 1796

1995 (n = 63)Bark 53.4 2.5 0.2 0.5 0.3 1.9 1548Threat calls (total) 7.5 27.9 0.2 0.4 0.2 2.7 1021

Puff 1.3 2.8 — — — 0.2 112Chuff — 1.0 — — — — 25Growl 6.1 16.4 0.2 0.2 0.0 1.1 633LITC 0.1 4.5 — 0.2 0.2 0.6 146GTC — 0.2 — — — 0.1 6FTC — 3.0 — — — 0.7 99

Submissive calls 0.0 0.2 — — — — 5Affiliative calls — — — — — — 0Unclassified 0.7 1.2 — 0.0 0.0 0.2 57N 1620 834 11 23 16 127 2631

Note: n is the number of sampling periods;N is the number of calls observed in each category. A value of 0.0 indicatesproportions <0.05%.

Table 3. Distribution (%) of calls made by adult male fur seals during the peak pupping periods of the 1994 and1995 seasons.



20 min after gaining a new territory (these calls did not occurduring observation periods dedicated to quantifying call us-age, and therefore are not represented in Table 3). The maledid not orient towards any animal in particular, and couldnot see any neighboring territorial males because of a highrock. When emitted in this context, FTCs appeared to becontagious, and neighboring males often began to call also.Territorial males also produced full-threat calls spontane-ously when a fight occurred in an adjacent territory.

Submissive calls

Submissive callThis is a high-pitched, strongly frequency modulated call,

typically around 1 s in duration (Table 6, Fig. 1h). Submis-sive calls were tonal, with dominant frequencies in the range600–1600 Hz. Calls were relatively complex, with numerouscall parts (Table 6) and pulses overlying the tonal structure(Fig. 1h). This was a high-intensity call and was often re-peated once or twice.

Context of submissive callsSubmissive calls were not emitted frequently by either fe-

males or males, although females produced them more often(3–4 vs. 0.1–0.2% of all vocalizations), and the majoritywere given between females (Tables 3, 4). The calls wereemitted by subordinate animals after an agonistic encounter

and were usually accompanied by open-mouth display or sub-missive posturing such as facing away (Fig. 2d). Smaller orapparently younger females used the submissive call whenthreatened by neighboring females. We also observed fe-males using the submissive call continuously as they at-tempted to move from one side of the colony to the other,particularly when moving to the tidal area in the morning.

Affiliative calls

Pup-attraction callsFemales typically made high-energy calls of moderate to

long duration (up to 2 s) when calling to their pups. Thecalls tended to be complex in structure and were typicallycomposed of an initial pulsed component followed by a tonalcomponent (fundamental frequency 0.7–1.0 kHz) that endedabruptly with a rapid decrease in frequency (Table 6, Fig. 1m).The rich harmonic structure in the latter part of the callsmade them sound distinctively high-pitched. PACs were highlyindividualistic (Phillips and Stirling 2000); some femalesproduced a pure, monotonic call with flat, parallel harmon-ics (Fig. 1n), while others had virtually no tonal componentswithin their calls (Fig. 1o).

Female-attraction callsThese calls also varied substantially among individuals

(Phillips and Stirling 2000) but, like PACs, were typically

© 2001 NRC Canada

426 Can. J. Zool. Vol. 79, 2001

Recipient of call

Call type Female Male Pup Juvenile Colony Unclassified N

1994 (n = 72)Bark — — 0.0 — — 0.1 12Threat calls (total) 19.4 5.8 0.6 0.5 0.0 4.3 5519

Puff 5.2 0.3 0.0 0.0 — 0.2 1047Chuff — — — — — — 0Growl 14.1 5.5 0.6 0.5 0.0 4.0 4445LITC — — — — — — 0GTC 0.1 0.0 0.0 0.0 — 0.0 26FTC — 0.0 — — — — 1

Submissive calls 2.7 0.1 — — — 0.0 518Affiliative calls 0.2 0.1 59.0 0.4 0.7 5.0 11405Unclassified 0.1 0.1 0.1 0.0 — 0.9 224N 4030 1108 10762 161 130 1840 18031

1995 (n = 63)Bark 0.4 — 0.0 0.0 — 0.1 80Threat calls (total) 25.5 10.7 1.2 1.2 — 0.9 6700

Puff 7.4 0.4 0.1 0.0 — 0.1 1358Chuff 0.0 — — — — — 1Growl 18.0 10.2 1.2 1.2 — 0.8 5332LITC — — — — — — 0GTC 0.0 0.0 — — — 0.0 9FTC — — — — — — 0

Submissive calls 3.5 0.4 — — — 0.0 670Affiliative calls — 0.1 55.4 0.3 — 0.1 9375Unclassified 0.2 0.0 0.1 0.0 — 0.1 55N 5030 1888 9646 253 0.0 193 17010

Note: n is the number of sampling periods;N is the number of calls observed in each category. A value of 0.0 indicatesproportions <0.05%.

Table 4. Distribution (%) of calls made by adult female fur seals during the peak pupping periods of the 1994 and1995 seasons.



pulsed at the start and ended with tonal elements that weresometimes strongly frequency-modulated (Table 6, Fig. 1p).Most FACs were high-energy, high-pitched calls, with afundamental frequency of 1.0–1.5 kHz, rich harmonic struc-ture, and many call parts (Table 6). Yearlings and juveniles(up to 3 years of age) that were not yet weaned also pro-duced a loud, high-pitched pulsed call with strong frequencymodulation (Fig.1q).

Context of affiliative callsPACs were given almost exclusively to pups or juveniles,

and were the most common call used by females at the peakof the pupping period (Table 4). On nine occasions we ob-served adult territorial males producing a call identical to thepulsed PAC, although the context was unknown (Table 3).Adult males were never heard to produce purely tonal callssuch as that illustrated in Fig. 1n or those described byPierson (1987: Fig. 4a).

PACs were produced under all circumstances in whichmothers call to their offspring, ranging from searching fortheir pups after returning from an offshore foraging trip(Figs. 2e, 2f ), to reuniting after being separated for minutesor hours on the beach, to lying in physical contact with theirpups. Females also gave PACs to the colony in general, espe-cially during the day or so immediately prior to parturition.We also observed mothers making these calls continuouslyduring parturition. Mother–pup reunion behavior is describedin Phillips (1998).

Pups up to about a week old appeared to give FACs indis-criminately to any passing female. After this, however, pupscalled primarily to their own mothers, either spontaneouslyor in response to their mother’s PAC. Yearlings that were notyet weaned continued to produce FACs, particularly whilesearching for their mothers within the colony. Calls of bothpups and yearlings were frequently accompanied by vigor-ous head shaking, particularly when their mother was return-ing from sea.

Vocal bouts between mothers and pups often occurredwhen the pair was resting, particularly in the early morning.These bouts were usually initiated by the female, who wouldgive several low-energy calls to her pup. If the pup awoke, itoften vocalized in response. The pair then called back andforth for up to a minute, at a rate of about 1–2 calls/s each,and nursing often followed. This type of restive calling mayserve to reinforce the bond between the mother and the pupand enable them to learn each other’s call (Phillips 1998).

In addition to resting bouts, a mother typically called toher pup when it left her side. She sometimes raised herselfup on her foreflippers to call, usually increasing the energyof the call as the pup moved away. The pup often remainedwhere it was and called back to its mother. At this point, sheeither went over and brought it back to where she had beensitting, or called once or twice more before going back tosleep, leaving the pup to sleep where it was or to wander offand play. To incite their pups to follow or remain close,mothers also called to them during their daily thermoregulatorymovements up or down the beach. If pups did not follow,mothers sometimes carried them through the colony by theirneck, which invariably resulted in loud, frequency-modulatedversions of pups’ calls.

Discussion

Barking and threat calls: graded signals?Barking occurs in all otariids (Table 7; for review see also

Miller 1991: 209–211). In most species, barks are tonalsounds that vary interspecifically in amplitude, duration, andfrequency characteristics. In sea lions and South African /Australian fur seals (Arctocephalus pusillus), barking is pro-duced with the mouth open, resulting in a long, loud “honk-ing” vocalization (Rand 1967), while other arctocephalinefur seals bark with the mouth nearly or completely closed,producing a sound like a soft, high-pitched “whimper” (Ta-ble 1; compare spectrograms in Peterson et al. 1968: Fig. 3;

© 2001 NRC Canada


(a) Adult males (34 of 6202 calls recorded).

Type of call

Bark Growl LITC Puff GTC FTC Snort Total

Sea lion (Otaria byronia)Adults 5 5 9 3 1 6 — 29Juveniles — — 1 — — — 2 3

Turkey vulture (Cathartes aura) 1 — — — — — — 1Crab (unidentified) — 1 — — — — — 1

(b) Adult females (66 of 44538 calls recorded).

Table 5. Summary of vocalizations directed towards other species by South American fur seals during the observation periods in bothyears (n = 220 h).

Type of call

Growl Puff Unclassified Total

Sea lion (O. byronia)Adults 3 5 2 10Juveniles 5 — — 5

Turkey vulture (C. aura) 36 3 1 40Peruvian gull (Larus belcheri) 2 — — 2Humboldt penguin (Spheniscus humboldtii) 7 1 — 8Crab (unidentified) 1 — — 1



©2

00

1N

RC

Ca

na

da

428C

an.J.

Zool.Vol.

79,2001

Bark (n = 20) Snort (n = 20) Puff (n = 20) Chuff (n = 12)

Mean SE Range Mean SE Range Mean SE Range Mean SE Range

Duration (ms) 46 2.7 20–72 115 5.8 81–170 83 3.8 62–112 132 17.4 72–247No. of parts 1 0 1 1 0 1 1 0 1 1 0 1Intercall interval (ms) 228 4.8 191–272 — — — 121 13.2 49–228 — — —Max. amplitude (V) 0.67 0.09 0.14–1.37 1.25 0.10 0.36–2.06 1.17 0.12 0.56–2.07 1.58 0.1 1.01–2.20

Growl (n = 20) LITC (n = 20) GTC (n = 20)

Mean SE Range Mean SE Range Mean SE Range

Duration (ms) 1922 562.0 169–10581 1292 96.5 534–1947 1177 84.5 628–1849No. of parts 1.6 0.3 1–6 4.5 0.9 1–13 10.8 0.7 5–17Initial frequency 365 23.4 158–593 623 42.6 339–915 613 56.7 457–915First peak frequency 721 22.9 598–1051 761 44.8 285–1123 782 41.1 573–1372Interpulse interval (ms) 11 0.9 5–18 17 2.8 5–49 55 5.8 30–128Max. amplitude (V) 0.44 0.04 0.21–0.75 0.83 0.09 0.18–1.43 0.90 0.08 0.45–1.89

FTC (n = 67) Submissive calls (n = 24) PAC (n = 300)a FAC (n = 260)a

Mean SE Range Mean SE Range Mean SE Range Mean SE Range

Duration (ms) 1405 40.9 759–2058 883 68.8 274–1504 1057 19.0 361–2231 600 13.9 181–1262No. of parts 8.2 0.4 2–18 5.5 0.5 2–11 2.7 0.1 1–7 4.7 0.2 1–15Harmonic interval 1150 36.4 550–1737 875 45.5 592–1618 905 11.4 530–1660 1030 18.5 410–1440Initial frequency 678 15.7 381–1016 808 29.4 586–1145 728 6.7 500–1310 1182 15.0 590–2000Max. frequency 1145 29.5 677–1780 1092 41.4 781–1618 944 11.2 560–1660 1470 21.3 750–2630End frequency 968 34.5 423–1780 880 28.9 625–1224 791 7.5 530–1380 1253 18.7 590–2190First peak frequency 1063 52.9 606–1586 912 29.3 642–1229 871 9.7 580–1510 1364 12.9 850–2260FM range — — — 339 47.4 156–742 353 27.6 90–1030 327 18.6 60–840FM period (ms) — — — 22 3.9 6–62 31 1.7 9–72 42 1.8 14–98

Note: All values are given in hertz, except number of parts, maximum amplitude, duration, intercall interval, interpulse interval and frequency modulation (FM) period.aData from Phillips and Stirling (2000).

Table 6. Summary of acoustic variables measured from each type of South American fur seal call.

J:\cjz\cjz79\cjz-03\Z00-219.vp

Wednesday, February 28, 2001 1:34:05 PM

Color profile: Disabled

Composite Default screen

©2

00

1N

RC

Ca

na

da

Phillips

andS

tirling429

Fig. 1. Spectrograms of vocalizations of South American fur seals (frequency bandwidth 39 Hz). (a) Bark. (b) Snort. (c) Puff. (d) Two chuffs. (e) Growl. (f ) Low-intensitythreat call (LITC). (g) Guttural threat call (GTC). (h) Two submissive calls, made by a retreating adult male after losing a fight with a territorial male. (i, j, andk) Full-threatcalls (FTCs) made by three different adult males. (l) FTC made by an adult female to a human observer. (m, n, ando) Pup-attraction calls (PACs) made by three differentmothers. (p) Female-attraction call (FAC) made by a pup. (q) FAC made by a yearling.

J:\cjz\cjz79\cjz-03\Z00-219.vp

Wednesday, February 28, 2001 1:34:20 PM

Color profile: Disabled

Composite Default screen

© 2001 NRC Canada

430 Can. J. Zool. Vol. 79, 2001

Peterson and Bartholomew 1969: Figs. 1, 2; Stirling andWarneke 1971: Fig. 4; Pierson 1987: Fig. 5a; Page et al.4:Fig. 3a). In contrast, the South American fur seal’s bark isa very brief, soft call with no harmonic structure, most simi-lar to that described for the northern fur seal,Callorhinusursinus(Lisitsyna 1973: Fig. 2-III). This likely explains whyTrillmich and Majluf (1981: 319) did not detect “anythingremotely resembling the barking of a sea lion (Zalophuscalifornianus) or a South African fur seal (A. pusillus).”Similar repetitive sounds have been reported for other mam-malian species and may have evolved as “an acoustic conse-quence of more active olfactory exploration” (Gould 1983:305).

Although the barking of South American fur seals differsstructurally to various degrees from that of other fur sealspecies, the call appears to have the same investigative func-tion. Adult males used it primarily towards adult females,especially while investigating potential sexual receptivity.Both adult males and females also appeared to use barkingtowards subordinate animals, perhaps to affirm their status orlocation, as suggested by Stirling and Warneke (1971). AtMacquarie Island, barks of male Antarctic, subantarctic, andNew Zealand fur seals show species-specific differences inacoustic structure, suggesting that barks have a role in spe-cies recognition and female mate choice in situations wherespecies’ ranges overlap (Page et al.3).

Snorts are subtle and difficult to detect and, not surpris-ingly, have only infrequently been reported in the literature(Table 7). Snorts appeared to be intermediate between bark-ing and puffing in both acoustic structure and apparent func-tion. In South American fur seals we found that snorts (andperhaps barking, when used between females) were used notonly to signal that more information was required about therecipient of the calls, but also to indicate a potential for in-creasing hostility. This is consistent with Winn and Schnei-der’s (1977: 831) observation that among pinnipeds, snortsmay have “universal significance in low-intensity warning.”

We suggest that the threat vocalizations used by SouthAmerican fur seals form two graded series of sounds, which,for convenience, we have termed respiratory calls and gut-tural calls. Respiratory calls are audible, noisy exhalations(sometimes followed by inhalations) that appear to form acontinuum of harsh, nontonal sounds ranging from soft, low-energy barking to loud, high-energy chuffs. We include bark-ing in this category because structurally it appears to be asoft, low-energy version of the puff (compare Figs. 1a and1c); however, inhalation sounds are not usually apparent dur-ing barking. Morton (1977) proposed that as sounds becomelouder and harsher, the signal indicates higher levels of hos-tility. In this study, the nontonal sounds made by the SouthAmerican fur seals became louder and more forceful as calls

intergraded from barking to snorting to puffing to chuffing,and the behaviors associated with them indicated increasingintensity of aggression and the potential for physical contact.Similarly, the guttural calls can also be thought of as a seriesof intergrading sounds that are based on the low-frequency,pulsed structure of a growl (Figs. 1e–1g). As the relative en-ergy levels, pulsation rate, and loudness of these calls in-creased, the hostility level appeared to rise.

Calls used in agonistic situations would be expected toshow gradation of structure and form. Since aggressive en-counters typically occur face to face or over short distances,graded vocalizations would allow fighting and display behav-ior to be more flexible, and animals would be more likely tomake use of the context of the situation to supplement thebehavioral message in the call (Smith 1977; Morton 1982;Miller 1991). For South American fur seals, context may in-clude the age, physical condition, and experience of the re-cipient. If adult males are able to recognize one another, assuggested by Stirling (1971a), Stirling and Warneke (1971),Roux and Jouventin (1987), and A.V. Phillips and I. Stirling(unpublished data and analyses), the identity of the senderwould provide important information to the recipient, partic-ularly if the latter has had previous experience interactingwith the sender. The importance of visual displays (such asfacial expression, body posture, and movement) made con-currently with vocal displays cannot be overlooked in theseinteractions (Miller 1975, 1991). For example, by makinghimself look bigger, a male may be able to convey a high-intensity threat with lower energy vocalizations. Also, subor-dinates may reduce the risk of injury by adjusting their threatvocalizations and posture in response to the behavior ofdominant animals.

Submissive callsIn contrast with the graded threat calls, discrete signals

are believed to indicate an evolutionary premium on lack ofambiguity (Altmann 1967). For example, in situations of ap-peasement and escape, it is critical for the retreating animalto make its message clear and thereby avoid further conflictand injury. The submissive calls of South American fur sealswere all typically long, loud, high-pitched, and directional,which would allow the sender to be readily located (Marler1955). Submissive calls amongArctocephalusspecies appearsimilar in acoustic structure and associated visual cues(Tables 1, 7; see Stirling 1971a: Fig. 5c; Miller 1975). Theseresults support Andersson’s (1980) proposal that submis-sive calls would be expected to be phylogenetically stablebecause of the potentially disastrous consequences to thesender of the signal if the meaning was misinterpreted.

The submissive calls of South American fur seals weresomewhat similar in structure to FACs used by yearlings and

Fig. 2. (a) Adult male barking to adult females; note the orientation of the head and vibrissae. (b) Adult male (at the left) puffing toan adult male; note the use of oblique staring. (c) Adult males exchanging high-intensity threat calls (GTCs and FTCs) during a terri-torial dispute. (d) Adult female (right) giving submissive calls while backing away from an adult male (left). (e) Mother calling to apup; note the body raised high on the foreflippers; the mouth is wide open. Her own pup is sleeping closest to her right foreflipper.( f ) The same mother and her pup engaging in naso-nasal inspection after the pup awoke and responded. Note the other pup movingaway from the right hind flipper of the adult female after having approached and been rejected by the female.

4B. Page, S.D. Goldsworthy, M.A. Hindell, and J. McKenzie. Individual vocal traits of mother and pup fur seals (Arctocephalusspp.).Submitted for publication.



© 2001 NRC Canada




2-year-olds: both were loud, high-pitched, and frequency-modulated (compare Figs. 1h and 1q). Juveniles would beexpected to produce higher pitched calls than adults becauseof physical limitations imposed by the size and musculatureof the vocal tract. However, the resemblance between sub-missive or appeasing calls and calls of juveniles has alsobeen reported for other mammals (e.g., Green 1975; Connerand Whitworth 1985). Morton (1977) suggested that submis-sive calls might have evolved as a vocal indication of smallsize, which might cause the recipient to recognize that thelevel of the perceived threat is lower and reduce its level ofhostility accordingly. Similar strategies appear to have evolvedin mountain sheep (Ovis canadensis), in which subordinaterams mimic the behavior of estrous ewes to avoid being at-tacked by dominant rams, while anestrous ewes mimic juve-nile behavior to reduce unwanted attention by rams (Geist1971).

Alternatively, in the South American fur seal, juvenilecalls may resemble submissive calls simply to reduce atten-tion from adult males. At Punta San Juan, submissive callsseem to cause dominant animals to reduce hostility or even

ignore the sender (consider the continuous use of these callsby subordinate females attempting to pass through the col-ony). Juvenile fur seals remaining at the breeding colony aretypically not yet weaned or have been weaned recently (Majluf1987), and maintain vocal contact with their mothers. Iftheir calls resemble submissive calls, juveniles might there-fore experience a reduced risk of agonistic or sexual inter-actions from territorial males.

Acoustically complex callsFull-threat and affiliative calls are also used in situations

in which the message conveyed by the call has to be unmis-takable. In the South American fur seal, FTCs indicate ahigh level of aggression and readiness to fight. PACs areonly used by mothers towards their young, whether to re-unite with the pup or maintain contact with it on the colony.Similarly, although pups are capable of producing a range ofsounds including puffs and growls, their high-pitched FACsare primarily used towards those females that they recognizeas their mother. Other studies onArctocephalusspp. havepostulated the same functions for these calls (Stirling 1971a;

© 2001 NRC Canada

432 Can. J. Zool. Vol. 79, 2001

Vocalization A. australis A. forsteri A. galapagoensis A. gazella

Bark Vaz-Ferreira 1956 Crawley and Wilson 1976; Miller 1971,1974, 1975; Page et al.3; Stirling1971a; Stirling and Warneke 1971

Bonner 1958, 1968; Page et al.3;Stirling and Warneke 1971

Snort Miller 1971, 1974; Stirling 1970Puff Trillmich and Majluf 1981

ChuffGrowl Trillmich and Majluf 1981;

Vaz-Ferreira 1971Miller 1971, 1974; Shaughnessy et al. 1988 Bonner 1958, 1968

LITC Trillmich and Majluf 1981;Vaz-Ferreira 1971

Crawley and Wilson 1976; Stirling 1970,1971a; Stirling and Warneke 1971

Stirling and Warneke 1971

GTC Vaz-Ferreira 1956 Crawley and Wilson 1976; Miller 1971;Stirling 1970, 1971a; Stirling andWarneke 1971

Stirling and Warneke 1971

FTC Trillmich and Majluf 1981 Crawley and Wilson 1976; Miller 1974;Page et al.3; Stirling 1971a; Stirlingand Warneke 1971

Page et al.3; Stirling andWarneke 1971

Submissive call Trillmich and Majluf 1981;Vaz-Ferreira 1956, 1971

Crawley and Wilson 1976; Miller 1975;Stirling 1971a; Stirling and Warneke 1971

Trillmich andMajluf 1981

PAC Majluf 1987; Trillmich andMajluf 1981; Vaz-Ferreira1956, 1971

Crawley and Wilson 1976; McNab andCrawley 1975; Page et al.4;Shaughnessy et al. 1988; Stirling 1971a,1971b; Stirling and Warneke 1971

Trillmich 1981 Bonner 1968; Page et al.4;Shaughnessy et al. 1988;Stirling and Warneke 1971

FAC Majluf 1987; Trillmich andMajluf 1981; Vaz-Ferreira1956, 1971

Crawley and Wilson 1976; McNab andCrawley 1975; Page et al.4; Stirling1971a; Stirling and Warneke 1971

Trillmich 1981 Bonner 1968; Page et al.4;Stirling and Warneke 1971

Othersb

Yelp Vaz-Ferreira 1971Female threat call Crawley and Wilson 1976; Stirling 1971a,

1971bMoan Crawley and Wilson 1976; Stirling 1971a

Note: The absence of a vocalization in the literature does not rule out the possibility that the vocalization occurs in the species.aSounds were recorded under water.bVocalizations that were not detected in this study.

Table 7. Summary of studies that refer to vocalizations made byArctocephalusspp. (excluding this study).



Stirling and Warneke 1971; Miller 1975, 1991; Trillmich1981; Pierson 1987; Roux and Jouventin 1987; see also Ta-ble 7).

The FTC of South American fur seals is comparable tothose illustrated for otherArctocephalusspecies (Stirling1971a: Figs. 5a, 5b; Stirling and Warneke 1971: Fig. 5d;Pierson 1987: Fig. 4; Roux and Jouventin 1987: Fig. 5; Pageet al.3: Fig. 3b; see also Tables 1, 7). The underlying com-pound form of the calls is apparent in most of the publishedspectrograms, even though individual variation is apparent inall of the species studied.

Roux and Jouventin (1987) suggested that compound FTCs(containing high-pitched tonal components) are emitted onlyby species that inhabit rugged terrain which limits the use ofvisual or olfactory cues for recognition among territorialmales. In contrast, they hypothesize that species which breedin more open habitat, such as the Galapagos (Arctocephalusgalapagoensis), Antarctic, and South African fur seals, wouldhave less use for FTCs. Whether such a trend exists is uncer-tain. The FTCs of Antarctic fur seals do contain tonal com-ponents (Page et al.3). According to Stirling and Warneke

(1971) and Miller (1991), South African fur seals do notproduce FTCs at all, but Rand (1967: 20) described a “deep-throated roar” and Caudron (1991) a “rugissement” (roar)produced by an adult male (although she stated that the con-text of this sound was difficult to determine in the captivesituation in her study). There are no published descriptionsof calls made by male Galapagos fur seals.

All fur seal females and pups produce affiliative calls (Ta-ble 7), and those of South American fur seals differ little inoverall structure from those of otherArctocephalusspecies.Harmonic structure appears to dominate the PACs of mostfur seals, although pulsation at the start of the calls is notuncommon (Stirling and Warneke 1971: Fig. 1; Trillmich1981: Figs. 2, 4; Pierson 1987: Fig. 8a; Roux and Jouventin1987: Fig. 5; Page et al.4). In all species, pup calls are typi-cally high-pitched and complex, as illustrated by the varietyof onomatopoeic terms found in the literature (e.g., bleat,wail, bawl, staccato call, whimper, baa; sources are as forTable 7). Calls of South American fur seal pups were similarto those illustrated for other species (Stirling and Warneke1971: Fig. 2; Trillmich 1981: Fig. 5; Pierson 1987: Fig. 8b;

© 2001 NRC Canada


A. philippii A. pusillus pusillus A. pusillus doriferus A. townsendi A. tropicalis

Norris and Watkins1971?a

Caudron 1991; Rand 1967 Stirling and Warneke 1971 Feldman 1993; Fleischer1978; Peterson et al.1968; Pierson 1987

Paulian 1964; Page et al.3;Rand 1967; Roux andJouventin 1987

Caudron 1991; Rand 1967Caudron 1991; Rand 1967 Feldman 1993; Fleischer

1978; Peterson et al. 1968Paulian 1964

Pierson 1987Norris and Watkins

1971Caudron 1991; Rand 1967 Feldman 1993; Fleischer

1978; Pierson 1987Paulian 1964; Shaughnessy et

al. 1988Stirling and Warneke 1971 Paulian 1964

Caudron 1991; Rand 1967 Stirling and Warneke 1971 Stirling and Warneke 1971 Stirling and Warneke 1971

Caudron 1991; Rand 1967 Feldman 1993; Peterson et al.1968; Pierson 1987

Paulian 1964; Page et al.3;Roux and Jouventin 1987

Rand 1967 Stirling and Warneke 1971 Stirling and Warneke 1971 Paulian 1964

Caudron 1991; Rand 1967 Stirling and Warneke 1971 Fleischer 1978; Peterson et al.1968; Pierson 1987

Paulian 1964; Page et al.4;Roux and Jouventin 1987;Shaughnessy et al. 1988

Cuadron 1991; Rand 1967 Stirling and Warneke 1971 Fleischer 1978; Pierson 1987 Paulian 1964; Page et al.4;Roux and Jouventin 1987

Caudron 1991; Rand 1967 Feldman 1993

Norris and Watkins1971

Rand 1967



Roux and Jouventin 1987: Fig. 6a; Page et al.4). Individualvariation in affiliative calls is presented in Phillips and Stirling(2000).

Although used in different circumstances, these three callshave several characteristics in common. They are all acousti-cally complex, typically consisting of both pulsed and tonalcomponents (compare Figs. 1i–1q). Unlike other threat vo-calizations and barking, they are typically long, loud callsand may be used over long distances as well as in close con-tact. Finally, all three appear to play an important role in in-dividual recognition, as has been reported for the mother andpup attraction calls (Phillips and Stirling 2000). FTCs alsoshow sufficient variation to enable recognition among indi-viduals (A.V. Phillips and I. Stirling, unpublished data andanalyses), as has been suggested in other studies of coloniallybreeding pinnipeds (Stirling 1971a; Shipley et al. 1981;Trillmich and Majluf 1981; Pierson 1987; Roux and Jouventin1987).

The structural similarity of full-threat and affiliative callsis likely related to the function of the calls. Calls used overlong distances, without the benefit of other sensory and con-textual cues, are typically of a discrete form in order to re-duce ambiguity (Marler 1976; Morton 1977, 1982; Miller1991). Furthermore, high-frequency sounds with harmonicstructure tend to be directional, and so are more common insounds used to attract the attention of the recipient or aid inthe location or detection of the sender (Marler 1955; Greenand Marler 1979). Also, complex calls allow large variationin both acoustic characteristics and syntactical organization,which would permit identity information to be encoded(Beecher 1989; Miller and Murray 1995). Finally, Morton(1977, 1982) and Miller (1991) further argue that these callsare all variants of a single display form, conveying strongarousal and disposition to interact.

The striking acoustic and functional similarities amongthese three calls used by animals at different stages of devel-opment lead us to speculate that the affiliative calls used bypups (FACs) are a developmental precursor to the “adult”PACs and FACs, especially as all have the same vital functionof individual recognition. Similar trends have been suggestedfor calls of northern elephant seals (Mirounga angustirostris)(Bartholomew and Collias 1962; Shipley et al. 1986). Insley(2000) found that calls made between northern fur sealmothers and pups are acoustically stable over time, enablingthem to recognize one another for several years. Thus, it isprobable that these acoustically complex calls represent thesame basic display, the context of which depends on the ani-mal’s ontogenetic state and sex and the social context (Miller1991). A structural comparison of these acoustically com-plex calls might provide more information on how calls de-velop with age. Sex-related differences in ontogeny are alsorequired, because in this study, adult female South Americanfur seals produced both PACs and FTCs. If females of otherspecies are also found to produce full-threat calls (see be-low), a structural comparison of these two types of callswould be informative.

Aggressive vocalizations and the problem ofinconsistency in the literature

Adult female South American fur seals are more aggres-sive towards conspecifics than any other species of fur seal

studied to date (Majluf 1987; 1992; Harcourt 1991, 1992).This is reflected in the range and frequency of agonisticcalls used by adult females in this study. Note that our in-vestigations of agonistic interactions were limited to vocalbehavior. Other studies have shown that visual displays suchas open-mouth threats are used more frequently than vocal-izations (e.g., Carey 1992), so caution should be used whencomparing our estimates of agonistic behavior with othersreported in the literature. Nonetheless, two hypotheses canbe put forward to explain the extent to which threat calls areused by adult female South American fur seals at Punta SanJuan compared with females of other populations and spe-cies. The first is that females of all fur seal species mighthave the vocal ability to produce the range of graded respi-ratory and guttural threat vocalizations that we observed inthis study, even though they are rarely used by most species.Fur seals at Punta San Juan might be subjected to a uniquecombination of factors not experienced by other species(Phillips 1998), resulting in the high level of aggression ob-served among adult female South American fur seals andsubsequent selective pressures on females to activate moreof the range of vocal threat signals than is necessary forfemales of other species. Second, the use of threat calls byfemales might simply have been underrepresented in the lit-erature. In many studies, threat calls have been named “male”calls, so there may have been a bias towards males whendescribing such calls. Also, most studies of the behavior offemale otariids have focused on reproductive behavior ormother–pup interactions, while the few studies that includedfemale aggression have tended not to incorporate vocal com-ponents. It is therefore possible that vocal threats made byfemales might simply have been overlooked or confusedwith other calls if they were used infrequently. Nevertheless,we suggest that the absence of information about aggressivevocalizations by females in the literature is more likely a re-flection of the difficulty of describing and quantifying be-haviors that are subtle or occur rarely (Lehner 1979), ratherthan their absence.

Studies of vocal behavior inArctocephalusand other otariidspecies have produced a multitude of terms for guttural andrespiratory threat calls (Table 1), leading to some confusionin the literature. For example, Pierson (1987: 87) describes a“boundary puff” as “a single sharp exhalation of air, whichresults in a harsh, puffing sound” and reports that it corre-sponds to the “male gutteral challenge” [sic] of Stirling andWarneke (1971). However, after examining Pierson’s spec-trogram (his Fig. 5b) and his description of the behavioralcontext of the call, we believe that it corresponds best towhat we term a chuff. This example illustrates the difficultyof using onomatopoeic descriptions to name and analyzegraded sounds (Marler 1976; Green and Marler 1979; Mor-ton 1982; Miller 1991).

We believe that there is a need to standardize nomencla-ture and acoustic analysis in research on pinniped vocal be-havior, as suggested by Stirling and Roux (1987) and Miller(1991). The availability of quantitative information on acousticstructure and call usage would allow interspecific comparisonsof behavior that could be used to address evolutionary andecological questions about the phylogeny of the variouspinniped taxa (e.g., Repenning et al. 1971; Berta and Wyss1994; Lento et al. 1997).

© 2001 NRC Canada

434 Can. J. Zool. Vol. 79, 2001



Our study is one of few that have attempted to quantifycall usage in otariids. This is important, because by verify-ing which calls were being used in which contexts, we re-duced the bias inherent in some of the call names (e.g.,female-attraction call). As is shown in Table 1, previousstudies have employed call nomenclature based on either theacoustic features of the call or its presumed function. Termsthat are based on functional or interpretive descriptions, suchas “female attraction” or “full threat” run the risk of beingused inappropriately, or used in an overgeneralized way(Lehner 1979; Miller 1991; Miller and Job 1992). Conversely,in order to compare across populations or species, purelyempirical terms such as “mother primary call” (Insley 1992)must be accompanied by a description of either the acousticcharacteristics or the context in which the call is most oftenused. Many studies, including this one, have also borrowedterms from common usage in the English language (e.g.,roar, growl, belch, whimper, bark). Although these do not di-rectly describe a function, there is a risk that in terms ofcontext, some readers might associate them with those ofmammals with which they are more familiar, such as bears,dogs, and cats.

Regardless of the system of nomenclature used, in studiessuch as these it is important to clearly define the terms usedfor specific vocalizations and behaviors, and not to adoptterms from the literature without confirming that they are ap-plicable to the species and circumstances under study. Care-ful examination of the literature is also required to ensurethat a single behavior is not given a multitude of names andascribed a multitude of functions. Finally, there is a need toquantify the response of the recipient, to better understandthe potential meanings of the calls, particularly those used inagonistic situations. As Paton (1986: 171) noted, “if the re-cipient—the individual to whom the signal is of most rele-vance—fails to make a distinction between [signals], wehave no evidence that the behavior patterns constitute differ-ent signals, no matter how different their respective formsmay be.”

Acknowledgements

We thank Patricia Majluf for inviting us to Punta San Juanand providing logistical support in the field as well as valuableadvice and access to unpublished material. Research effortsat Punta San Juan are supported by Wildlife ConservationInternational, Pesca-Peru, Hierro-Peru, and UniversidadPeruanaCayetano Heredia. Special thanks are extended toGabriella Battistini and Tanya Luszcz for their assistance inthe field. Logistical support for this study was provided bythe Canadian Wildlife Service (CWS), Prairie and NorthernRegion. Special thanks are extended to Nick Lunn, WendyCalvert, and Dennis Andriashek for assistance with field andlaboratory equipment and helpful comments on the manu-script, and especially to Sam Barry for extensive statisticalhelp. The manuscript was improved by discussion and com-ments from Jan Murie, Ted Miller, Steve Insley, and twoanonymous reviewers. We thank the late Luis Baptista andhis staff at the California Academy of Sciences for addi-tional access to SIGNAL. Financial support for this study wasprovided by an operating grant to I.S. from the Natural Sci-ences and Engineering Council of Canada (NSERC), and an

NSERC postgraduate scholarship, the Walter H. John Grad-uate Fellowship, and a Department of Biological SciencesGraduate Teaching Assistantship to A.V.P.

References

Adams, D.C., and Anthony, C.D. 1996. Using randomization tech-niques to analyse behavioural data. Anim. Behav.51: 733–738.

Altmann, J. 1974. Observational study of behavior: sampling meth-ods. Behaviour,49: 227–267.

Altmann, S.A. 1967. The structure of primate social communica-tion. In Social communication among primates.Edited byS.A.Altmann. University of Chicago Press, Chicago. pp. 325–362.

Andersson, M. 1980. Why are there so many threat displays? J.Theor. Biol. 86: 773–781.

Bartholomew, G.A., and Collias, N.E. 1962. The role of vocaliza-tion in the social behavior of the northern elephant seal. Anim.Behav.10: 7–14.

Beecher, M.D. 1989. Signalling systems for individual recognition:an information theory approach. Anim. Behav.38: 248–261.

Berta, A., and Deméré, T.A. 1986.Callorhinus gilmorei n. sp.,(Carnivora: Otariidae) from the San Diego Formation (Blancan)and its implications for otariid phylogeny. Trans. San DiegoSoc. Nat. Hist.21: 111–126.

Berta, A., and Wyss, A.R. 1994. Pinniped phylogeny. Proc. SanDiego Soc. Nat. Hist.29: 33–56.

Bonner, W.N. 1958. Notes on the southern fur seal in South Geor-gia. Proc. Zool. Soc. Lond.130: 241–252.

Bonner, W.N. 1968. The fur seal of South Georgia. Br. Antarct.Surv. Sci. Rep. No. 56.

Carey, P.W. 1992. Agonistic behaviour in female New Zealand furseals,Arctocephalus forsteri. Ethology,92: 70–80.

Caudron, A. 1991. Étude comportementale d’un groupe d’otaries àfourrure d’Afrique du Sud (Arctocephalus pusillus pusillusSchreber1776) en captivité : éthogramme, budget d’activité, structuresociale. M.Sc. thesis, Université de Liège, Liège, Belgium.

Conner, D.A., and Whitworth, M.R. 1985. The ontogeny of vocalcommunication in the pika. J. Mammal.66: 756–763.

Crawley, M.C., and Wilson, G.J. 1976. The natural history and be-haviour of the New Zealand fur seal (Arctocephalus forsteri).Tuatara,22: 1–29.

Crowley, P.H. 1992. Resampling methods for computation-intensivedata analysis in ecology and evolution. Annu. Rev. Ecol. Syst.23: 405–477.

Feldman, H.N. 1993. Vocalizations in a juvenile female Guadalupefur seal.In Proceedings of the 10th Biennial Conference on theBiology of Marine Mammals, Galveston, Texas, 11–15 November1993. Society for Marine Mammalogy, Lawrence, Kans. p. 47.

Fleischer, L.A. 1978. The distribution, abundance, and populationcharacteristics of the Guadalupe fur seal,Arctocephalus townsendi(Merriam 1897). M.Sc. thesis, University of Washington, Seattle.

Geist, V. 1971. Mountain sheep: a study in behavior and evolution.University of Chicago Press, Chicago.

Goldsworthy, S.D., Boness, D., and Fleischer, R. 1999. Mate-choiceamong sympatric fur seals: female preference for con-phenotypicmales. Behav. Ecol. Microbiol.45: 253–267.

Gould, E. 1983. Mechanisms of mammalian auditory communica-tion. Spec. Publ. Am. Soc. Mammal. No. 7. pp. 265–342.

Green, S. 1975. Variation of vocal pattern with social situation inthe Japanese monkey (Macaca fuscata): a field study.In Primatebehaviour: developments in field and laboratory research. Vol. 4.Edited byL.A. Rosenblum. Academic Press, New York. pp. 1–102.

Green, S., and Marler, P. 1979. The analysis of animal communi-cation. In Handbook of behavioral neurobiology. Vol. 3. Social

© 2001 NRC Canada




© 2001 NRC Canada

436 Can. J. Zool. Vol. 79, 2001

behaviour and communication.Edited by P. Marler and J.G.Vandenbergh. Plenum Press, New York. pp. 73–158.

Harcourt, R. 1991. Maternal aggression in the South American furseal in Peru. Can. J. Zool.70: 320–325.

Harcourt, R. 1992. Factors affecting early mortality in the SouthAmerican fur seal (Arctocephalus australis) in Peru: density-related effects and predation. J. Zool. (Lond.),226: 259–270.

Insley, S.J. 1992. Mother–offspring separation and acoustic stereotypy:a comparison of call morphology in two species of pinnipeds.Behaviour,120: 103–121.

Insley, S.J. 2000. Long-term vocal recognition in the northern furseal. Nature (Lond.),406: 404–405.

Lehner, P.N. 1979. Handbook of ethological methods. Garland STPMPress, New York.

Lento, G.M., Haddon, M., Chambers, G.K., and Baker, C.S. 1997.Genetic variation of southern hemisphere fur seals,Arctocephalusspp.: investigation of population structure and species identity.J. Hered.88: 202–208.

Lisitsyna, T.Yu. 1973. Behaviour and acoustic signals of the northernfur seal (Callorhinus ursinus) at lairs. Zool. Zh.52: 1220–1228.

Majluf, P. 1987. Reproductive ecology of female South Americanfur seals at Punta San Juan, Peru. Ph.D. thesis, University ofCambridge, Cambridge, U.K.

Majluf, P. 1992. Timing of births and juvenile mortality in theSouth American fur seal in Peru. J. Zool. (Lond.),227: 367–383.

Majluf, P., and Goebel, M.E. 1992. The capture and handling offemale South American fur seals and their pups. Mar. Mamm.Sci. 8: 187–190.

Marler, P. 1955. Characteristics of some animal calls. Nature (Lond.),176: 6–8.

Marler, P. 1976. Social organization, communication and gradedsignals: the chimpanzee and the gorilla.In Growing points inethology.Edited byP.P.G. Bateson and R.A. Hinde. CambridgeUniversity Press, Cambridge. pp. 239–280.

McNab, A.G., and Crawley, M.C. 1975. Mother and pup behaviourof the New Zealand fur seal,Arctocephalus forsteri(Lesson).Mauri Ora,3: 77–88.

Miller, E.H. 1971. Social and thermoregulatory behaviour of theNew Zealand fur seal,Arctocephalus forsteri(Lesson, 1828).M.Sc. thesis, University of Canterbury, Canterbury, New Zealand.

Miller, E.H. 1974. Social behaviour between adult male and femaleNew Zealand fur seals,Arctocephalus forsteri(Lesson) duringthe breeding season. Aust. J. Zool.22: 155–173.

Miller, E.H. 1975. A comparative study of facial expressions oftwo species of pinnipeds. Behaviour,53: 268–284.

Miller, E.H. 1991. Communication in pinnipeds, with special refer-ence to non-acoustic signalling.In The behaviour of pinnipeds.Edited byD. Renouf. Chapman and Hall, London. pp. 128–235.

Miller, E.H., and Job, D.A. 1992. Airborne acoustic communica-tion in the Hawaiian monk seal,Monachus schauinslandi. InMarine mammal sensory systems.Edited byJ.A. Thomas, R.A.Kastelein, and A.E. Supin. Plenum Press, New York. pp. 485–531.

Miller, E.H., and Murray, A.V. 1995. Structure, complexity, andorganization of vocalizations in harp sealPhoca groenlandicapups. In Sensory systems of aquatic mammals.Edited byR.A.Kastelein, J.A. Thomas, and P.E. Nachtigall. DeSpil Publishers,Woerden, the Netherlands. pp. 237–264.

Moore, P.W., and Schusterman, R.J. 1987. Audiometric assessmentof northern fur seals,Callorhinus ursinus. Mar. Mamm. Sci.3:31–53.

Morton, E.S. 1977. On the occurrence and significance of motiva-

tion-structural rules in some bird and mammal sounds. Am. Nat.111: 855–869.

Morton, E.S. 1982. Grading, discreteness, redundancy, and motivation–structural rules.In Acoustic communication in birds. Vol. 1.Production, perception, and design features of sounds.Edited byD.E. Kroodsma and E.H. Miller. Academic Press, New York.pp. 183–212.

Norris, K.S., and Watkins, W.A. 1971. Underwater sounds ofArctocephalus philippii, the Juan Fernández fur seal. Antarct.Res. Ser.18: 169–171.

Paton, D. 1986. Communication by agonistic displays: II. Per-ceived information and the definition of agonistic displays.Behaviour,99: 157–175.

Paulian, P. 1964. Contribution à l’étude de l’Otarie de l’île Amster-dam. Mammalia,28(Suppl. 1): 1–146.

Peters, G., and Wozencraft, W.C. 1989. Acoustic communicationby fissiped carnivores.In Carnivore behavior, ecology, and evolu-tion. Edited byJ.L. Gittleman. Cornell University Press, Ithaca,N.Y. pp. 14–56.

Peterson, R.S., and Bartholomew, G.A. 1969. Airborne vocal com-munication in the California sea lion,Zalophus californianus.Anim. Behav.17: 17–24.

Peterson, R.S., Hubbs, C.L., Gentry, R.L., and DeLong, R.L. 1968.The Guadalupe fur seal: habitat, behaviour, population size, andfield identification. J. Mammal.49: 665–675.

Phillips, A.V. 1998. Vocal communication and mother–pup inter-actions in the South American fur seal,Arctocephalus australis.M.Sc. thesis. University of Alberta, Edmonton.

Phillips, A.V., and Stirling, I. 2000. Vocal individuality in motherand pup South American fur seals,Arctocephalus australis. Mar.Mamm. Sci.16: 592–616.

Pierson, M.O. 1987. Breeding behaviour of the Guadalupe fur seal,Arctocephalus townsendi. NOAA Tech. Rep. NMFS No. 51.pp. 83–94.

Rand, R.W. 1967. The Cape fur seal (Arctocephalus pusillus). 3.General behaviour on land and at sea. Invest. Rep. Div. SeaFish. S. Afr.60: 1–39.

Renouf, D. 1991. Sensory reception and processing in Phocidaeand Otariidae.In The behaviour of pinnipeds.Edited by D.Renouf. Chapman and Hall, London. pp. 345–394.

Repenning, C.A., Peterson, R.S., and Hubbs, C.L. 1971. Contribu-tions to the systematics of the southern fur seals, with particularreference to the Juan Fernàndez and Guadalupe species. Antarct.Res. Ser.18: 1–34.

Roux, J.-P., and Jouventin, P. 1987. Behavioral cues to individualrecognition in the subantarctic fur seal,Arctocephalus tropicalis.NOAA Tech. Rep. NMFS No. 51. pp. 95–102.

Shaughnessy, P.D., Shaughnessy, G.L., and Fletcher, L. 1988. Re-covery of the fur seal population at Macquarie Island. Pap. Proc.R. Soc. Tasman.122: 177–187.

Shipley, C., Hines, M., and Buchwald, J.S. 1981. Individual differencesin threat calls of northern elephant seal bulls. Anim. Behav.29:12–19.

Shipley, C., Hines, M., and Buchwald, J.S. 1986. Vocalizations ofnorthern elephant seal bulls: development of adult call charac-teristics during puberty. J. Mammal.67: 526–536.

Smith, W.J. 1977. The behaviour of communicating. Harvard Uni-versity Press, Cambridge, Mass.

Stirling, I. 1970. Observations on the behaviour of the New Zea-land fur seal (Arctocephalus forsteri). J. Mammal.51: 766–778.

Stirling, I. 1971a. Studies on the behaviour of the South Australianfur seal,Arctocephalus forsteri(Lesson). I. Annual cycle, pos-



© 2001 NRC Canada


tures and calls, and adult males during the breeding season.Aust. J. Zool.19: 243–266.

Stirling, I. 1971b. Studies on the behaviour of the South Australianfur seal,Arctocephalus forsteri(Lesson). II. Adult females andpups. Aust. J. Zool.19: 267–273.

Stirling, I., and Roux, J.-P. 1987. Fur seal vocalizations. NOAATech. Rep. NMFS No. 51. pp. 201–202.

Stirling, I., and Warneke, R.M. 1971. Implications of a comparisonof the airborne vocalizations and some aspects of the behaviourof the two Australian fur seals,Arctocephalusspp., on the evo-lution and present taxonomy of the genus. Aust. J. Zool.19:227–241.

Tollu, B. 1982. Les communications chez les otaries à fourrure. J.Psychol.1982: 99–123.

Trillmich, F. 1981. Mutual mother–pup recognition in Galápagosfur seals and sea lions: cues used and functional significance.Behaviour,78: 21–42.

Trillmich, F., and Majluf, P. 1981. First observations on colonystructure, behaviour, and vocal repertoire of the South American

fur seal (Arctocephalus australisZimmermann, 1783) in Peru.Z. Saeugetierkd.46: 310–322.

Vaz-Ferreira, R. 1956. Etología terrestre deArctocephalus australis(Zimmermann) (“lobo fino”) en las Islas Uruguayas. Trabajossobre islas de lobos y lobos marinos No. 2, Ministerio de Industriasy Trabajo Servicio Oceanografico y de Pesca, Montevideo, Uruguay.

Vaz-Ferreira, R. 1971. Repertorio vocal aero y sus funciones socialesen Arctocephalus australis(Zimmerman) y Otaria byronia(Blainville). Bol. Soc. Zool. Uruguay,1: 62–65.

Winn, H.E., and Schneider, J. 1977. Communication in sireniens,sea otters, and pinnipeds.In How animals communicate.Editedby T.A. Sebeok. Indiana University Press, Bloomington, Ind.pp. 809–840.

Wynen, L.P., Goldsworthy, S.D., Guinet, C., Bester, M.N., Boyd,I.L., Gjertz, I., Hofmeyr, G.J.G., White, R.G., and Slade, R.2000. Postsealing genetic variation and population structure oftwo species of fur seal (Arctocephalus gazellaandA. tropicalis).Mol. Ecol. 9: 299–314.



Documents

Vocal repertoire of South American fur seals, Arctocephalus australis : structure, function, and context