Docstoc

Nicholson_ Teri Underwater Vocalization and Subsequent Social

Document Sample
Nicholson_ Teri Underwater Vocalization and Subsequent Social Powered By Docstoc
					 Nicholson, Teri: Underwater Vocalization and Subsequent Social Beh...hardsi) Near Hopkins Marine Station, Pacific Grove, CA, MBARI 1997




Underwater Vocalization and Subsequent Social
Behavior of the Harbor Seal (Phoca vitulina richardsi)
Near Hopkins Marine Station, Pacific Grove, CA
Teri Nicholson, Moss Landing Marine Laboratories
Mentors: Khosrow Lashkari, Dave Mellinger
Summer 1997
Keywords: behavior, seals, acoustics


ABSTRACT
Underwater observations of vocalizations and subsequent social behavior of individually identified adult
male harbor seals were recorded at Hopkins Marine Station from November 1996 to May 1997. When
males vocalized, nearby males either ignored or responded to vocalizations by approaching and posturing
submissively nose to nose with the vocalizer. We defined this social behavior as attending. During this
study, we analyzed underwater vocalizations of three attended and six non-attended adult males to (1)
identify reliable acoustic parameters for individual comparison, and (2) determine whether vocalizations
of males routinely attended during acoustic displays differed from males that are ignored. Underwater
vocalizations were divided into three primary call types: preroar, step and roar. We compared roar sound
spectra and mean first formant frequency (Fo) of the same male at constant depth and distance. but
opposite orientation to the hydrophone to investigate effects of multipath sound propagation in shallow
water, and determine the reliability of spectral measurements. Spectra differed significantly with respect
to orientation (t = 3.0, p = 0.01, df = 53), but mean first formant frequency (F0) remained consistent (t =
0.88, p = 0.40. df = 9). Mean frequency (Fo), therefore, was a reliable measure of sound frequency.
Duration of preroar, step, and roars, and mean frequency (Fo) of roars were compared between
non-attended and attended males. Attended males produced longer (3.3 s, SE 0.2), and lower frequency
(186.1 Hz, SE 13.0) roars than non-attended seals (2.0 s, SE 0.1; 222.0 Hz, SE 8.0, t = 2.8, p = 0.03, df =
7), indicating that seals interpret social status from acoustic cues. This communication of fitness or social
status is relevant to mate selection as a means to mediate dominance hierarchies, maintain underwater
territories, or as a direct advertisement to females. Future studies will focus on distinguishing among
mating strategies by improving acoustic monitoring capabilities.
INTRODUCTION


 http://www.mbari.org/interns/projects/papers/97Nicholson.html (1 of 6) [11/3/1999 10:03:11 AM]
 Nicholson, Teri: Underwater Vocalization and Subsequent Social Beh...hardsi) Near Hopkins Marine Station, Pacific Grove, CA, MBARI 1997

Harbor seal (Phoca vitulina richardsi) mating system is polygynous (Bartholomew 1970, Le Boeuf
1991). Polygynous mating systems are characterized by variability in reproductive success that depends
upon establishment of dominance relationships among males competing for access to estrous females. In
reproductive systems of pinnipeds that mate on shore; such as otariids, elephant seals (Mirounga
angustirostris and Mirounga leonina), and grey seals (Haliochoerus grypus); dominance relationships
are established by resource or female defense strategies (Bartholomew 1970, Le Boeuf 1991). Dominant
males obtain breeding privileges by controlling access to breeding habitat required by females. Harbor
seals, however, mate offshore where males are unable to compete for these resources. Alternative mating
strategies include establishment of leks, dominance hierarchies, or underwater territories (Sullivan 1981,
Le Boeuf 1991, Hanggi and Schusterman 1994, Thompson et al. 1994, Van Parijs et al. 1996).
During the breeding season, male harbor seals are thought to maintain leks, dominance hierarchies, or
underwater territories by performing underwater acoustic displays (Hauggi and Schusterman 1994,
Coltruan 1996, Van Parijs et al. 1996). Similar hypotheses regarding the function of vocalizations in
mating strategy have been described for other phocids that mate underwater, such as the bearded seal
(Erignathus barbatus; Ray et al. 1969, Cleator et al. 1989, Ajmi 1996), harp seal (Pagophilus
groenlandicus; Mohl et al. 1975), leopard seal (Hydrurga leptonyx; Rogers et al. 1996), ringed seal
(Phoca hispida; Stirling 1973, Kunnasranta et al. 1996), and walrus (Odobenus rosmarus; Ray and
Watkins 1975, Stirling et al. 1987). How females choose a mate, however, is poorly understood because
reproductive behavior has not been directly observed.
Males may communicate their fitness to females during underwater vocalizations. Hanggi and
Schusterman (1994) measured differences in sound frequency characteristics of underwater vocalizations
among individual males. These characteristics may correspond with a male’s physical attributes, such as
size or health, which are relevant to mate selection. Seals were heard vocalizing during this study, but
underwater behaviors associated with the vocalizations were not observed. Study of the associated
behavior and social context, however, is necessary to determine the function of vocalizations with respect
to mating and social strategies.
Harbor seals exhibit fidelity to haul-out areas while resting ashore (Pitcher and McAllister 1981. Harvey
1987). I have identified more than 350 resident harbor seals near Hopkins Marine Station in Monterey
Bay. California from 1994 to 1996. Individuals are distinguished by unique pelage markings and can be
observed onshore or underwater at close range year round. From September 1996 to May 1997
underwater vocalizations and social behavior of thirty adult males were recorded using a housed Sony
TR8 1 video camera and attached hydrophone.
When males vocalized, nearby males either ignored or responded to vocalizations by approaching or
attending. Attending seals postured submissively by positioning themselves nose to nose with the
vocalizer during a call. Five (5) older adult males within this community were attended during more than
eighty-five (85) percent of vocalization bouts. Mean number of attendants was 2.0 (n= 163, SD 1.2, max
6). Younger adult males (21) were attended during less than five (5) percent of vocalization bouts, and
never attended by more than one other adult male at a time. These underwater observations of social
behavior associated with underwater vocalizations indicated that males may maintain social hierarchies
by communicating social status while performing vocalization displays (Nicholson, unpubl. data).
The objectives of this study were to (1) provide a preliminary description of vocal characteristics of
individuals within this community with an emphasis on determining reliable parameters to compare


 http://www.mbari.org/interns/projects/papers/97Nicholson.html (2 of 6) [11/3/1999 10:03:11 AM]
 Nicholson, Teri: Underwater Vocalization and Subsequent Social Beh...hardsi) Near Hopkins Marine Station, Pacific Grove, CA, MBARI 1997

differences among individuals, and (2) correlate these differences with observed social behavior during
and subsequent to underwater vocalizations by determining how vocalization characteristics of adult
males routinely attended during acoustic displays differ from males that are ignored.
MATERIALS AND METHODS
Underwater vocalizations of individually identified adult males were obtained from twenty (20) hours of
hi-8 video/audio tapes recorded from November 1996 to May 1997 at Hopkins Marine Station, using a
housed Sony TR81 video camera and hydrophone. Individuals were recognized by unique pelage
patterns. We generated spectrograms of each vocalization using Canary (frame size 50% overlap,
hamming window, Fourier band filter width 175 Hz). Sampling rate was 22.05 kHz. Analyses were
limited to nine males which had been recorded more than fifteen times (n>15).
We also categorized vocalizations of the most frequently recorded adult male by his orientation to the
hydrophone at constant depth and distance to determine effects of orientation on sound propagation or
sound spectral characteristics recorded by the hydrophone, and investigate the reliability of spectral
measurements. Two orientation categories were defined as (1) the seal facing away from the hydrophone
by more than 120 degrees or (2) toward the hydrophone by less than 60 degrees. Spectral characteristics
were compared by computing the cross correlation between each pair of spectra (Zar 1984). We
compared underwater vocalizations recorded from similar orientations and opposite orientations by
comparing mean cross correlations of vocalizations recorded from (1) within each orientation category
and (2) between orientation categories using a Student’s t test.
Individuals adult males were classified into two distinct behavior categories: attended and non-attended
seals. Attended males (n = 3) attracted nearby males, positioning themselves nose to nose with the
vocalizer, during more than eighty-five percent of underwater vocalization bouts. Non-attended males (n
= 6) vocalized alone during more than ninety-five percent of vocalization bouts.
Vocalizations were divided into three call types: preroar, step, and roar (fig. 1). Grunts, groans and
creaks were also recorded, but not frequently enough for analysis. We measured the duration of call types
for each vocalization, and sound frequency of roars by calculating the mean first formant frequency (Fo)
or lowest energy band of the vocalization.
We compared mean first formant frequencies (Fo) of the same seal in opposite orientations but constant
depth and distance from the hydrophone using a Student’s t test to determine the reliability of this
frequency measure. Finally, we tested for differences in mean duration of call types (preroar, step, and
roar) and sound frequency (Fo) of roars between attended and non-attended seals using a Student’s t test.
Mean frequency (Fo) of roars were compared among all individuals using a single factor ANOVA and
Tukey multiple comparison test.


RESULTS
Sound propagation--Orientation of seals with respect to the hydrophone significantly affected sound
spectra of underwater roars (t = 3.0, p = 0.01, df = 53). Specifically, peak frequencies and intensity levels
(dBs) between 50 and 3000 Hz varied with respect to whether the seal faced toward or away from the
hydrophone while vocalizing (Table 1). Orientation, however, did not significantly affect mean first


 http://www.mbari.org/interns/projects/papers/97Nicholson.html (3 of 6) [11/3/1999 10:03:11 AM]
 Nicholson, Teri: Underwater Vocalization and Subsequent Social Beh...hardsi) Near Hopkins Marine Station, Pacific Grove, CA, MBARI 1997

formant frequency (F0) (t =0.88, p = 0.40, df= 9, Table 2).

Characteristics of call types--Preroar was the lowest and narrowest frequency (<250 Hz) vocalization,
and preceded steps and roars in call sequence (fig. 1). Mean duration of preroar was 2.3 s, ranging from
0.1 to 6.3 s. Roars were flat spectrum calls with primary frequency range from 50 to 3000 Hz, mean and
maximum duration of 2.6 s and 6.0 s, respectively. This call was non-harmonic, but characterized by
broad frequency bands resembling formants. Steps resembled roars with frequency range less than 1000
Hz. Among the three call types, seals used step calls least frequently (mean 1.2 s, max 6.2 s).
Relative duration of call types--Attended seals spent proportionally more time roaring (0.52 ± 0.04)
while vocalizing than non-attended seals (0 33 ± 0 03, t = 7.6, p = 0.000, df = 7). Non-attended seals
spent proportionally more time producing preroar (0.56 ± 0.03) and step calls (0.43 ± 0.08) than attended
seals (0.39 ± 001, 0 23 ± 0.09; t = 10.2, 3.6; p = 0.00, 0.01; df = 7; Table 3).
Roar duration and frequency--Attended seals produced roars of longer duration (3.3 ± 0.2 s) and lower
mean frequency (Fo, 186.1 ± 13.0 Hz) than non-attended seals (2.0 ± 0.1 s 222.0 ± 8.0 Hz; t = 6.2, 2.8; p
= 0.00, 0.03; df = 7; fig. 2, fig. 3). These differences in roar characteristics between the two groups are
apparent when mean frequency (Fo) and duration are plotted together (fig. 4). Mean frequency (Fe)
among individual seals (2, 4, 6, 7, 10, 12, 15, 16, and 17) was also significantly different (f= 10.9,
p<0.0001), but only for 13 of 36 paired comparisons.


DISCUSSION
Characterization of underwater sounds produced by seals in a near shore, shallow water environment is
complicated by multipath sound propagation. Specifically, as sound travels from seal to hydrophone, the
wave path propagates by bouncing off the surface or bottom of the ocean, or both as in reverberation (fig.
5). Consequently, underwater vocalizations received by the hydrophone are distorted. Sound waves
bouncing off the surface reverse phase and destructively interfere or mask sound at specific frequencies
while sound waves deflecting off the bottom constructively interfere or increase energy or loudness.
Multipath sound propagation, therefore, may have contributed to the significant sound spectral
differences we observed in an individual’s underwater vocalizations with respect to orientation. A seal’s
distance or depth may also affect sound propagation as well as environmental factors such as substrate
composition, topography, and surface waves.
Harbor seals produce calls by contracting and expanding throat, thoracic and chest area, creating a
turbulent flow of air through a constriction in their airway. The resulting sound is an aperiodic, flat
spectrum sound with broad frequency bands called formants. This contrasts with periodic sound, or
harmonics produced by vibration of vocal cords. The first formant (Fo) was the most reliable and
conservative measure of sound frequency considering measured variation in sound spectra related to
propagation in shallow water (Tables 1, 2). Although we did not measure frequency differences between
each pair of adult males, significant differences among seals indicated that individual identification may
be possible with better quality underwater recordings from a stationary, bottom mounted hydrophone
beyond the surf zone and shallow water environment.
Regardless of the limitations associated with recording in shallow water, this is the first study to

 http://www.mbari.org/interns/projects/papers/97Nicholson.html (4 of 6) [11/3/1999 10:03:11 AM]
 Nicholson, Teri: Underwater Vocalization and Subsequent Social Beh...hardsi) Near Hopkins Marine Station, Pacific Grove, CA, MBARI 1997

document a relationship between underwater vocalizations and social behavior of harbor seals.
Characteristics of vocalizations are related to health, size and experience of the individual. Specifically,
sound production apparatus affects acoustic output such that animals with long vocal tract lengths
produce vocalizations with low formant frequencies (Fant 1960). Large, healthy experienced males,
therefore, may be able to produce longer, lower frequency vocalizations than other males. When males
produce these calls nearby males responded by approaching and submissively posturing nose to nose
with the roaring seal, indicating that seals interpret social status from acoustic cues. This communication
of fitness or social status is relevant to mate selection as a means to mediate dominance hierarchies
among males maintain underwater territories, or as a direct advertisement to females.
Future studies will focus on distinguishing among these mating strategies by improving acoustic
monitoring capabilities. Specifically, we will record harbor seals in deeper water to avoid problems with
multipath sound propagation and obtain more reliable measures of sound frequency of individuals.
Further, we will extend the area in which we record to monitor the frequency, distribution and social
interactions among these individuals. These acoustic monitoring capabilities coupled with extensive
knowledge of individuals within this community are well suited to address questions regarding the role
of sound in the reproductive strategies of harbors seals; specifically, how seals develop sound production
and use it for communication, and how this relates to innate selection and success.


REFERENCES
Ajmi, AR. 1996. Underwater bioacoustic analysis of bearded seal behavior off Barrow, Alaska. MS
thesis, Univ. of Alaska. 115pp.
Allen, SG. 1988a. Movement and activity patterns of harbor seals at the Point Reyes Peninsula,
California. MS thesis, UC Berkeley, 70 pp.
Bartholomew, GA. 1970. A model for the evolution of pinniped polygyny. Evolution, 24:546-559.
Brown, RF, and BR Mate. 1983. Abundance, movement, and feeding habits of harbor seals, Phoca
vitulina, at Netarts and Tillamook Bays, Oregon. Fish. Bull., 81:291-301.
Coltman, DW. 1996. Phenotype and mating success of male harbour seals, Phoca vitulina, at Sable
island, Nova Scotia. Ph.D. thesis, Dalhousie Univ. 185 pp.
Cleator,HJ, I Stirling, and TG Smith. 1989. Underwater vocalizations of the bearded seal (Erignathus
barbatus). Can. J. Zool., 67: 1900-1910.
Fant, G. 1960. Acoustic Theory of Speech Production. Mouton, The Hague.
Hanggi, EB, and RI Schusterman. 1994. Underwater acoustic displays and individual variation in male
harbor seals, Phoca vitulina. Anim. Beh., 48:1275-1283.
Harvey, iT. 1987. Population dynamics, annual food consumption, movement, and dive behaviors of
harbor seals, Phoca vitulina richardsi, in Oregon. Unpubl. Ph.D. Dissert., Oregon State Univ., Corvallis,
177 pp.
Huber, 1{R, SJ Jeffries, RF Brown, and RI Delong. 1995. Correcting minimum population counts to


 http://www.mbari.org/interns/projects/papers/97Nicholson.html (5 of 6) [11/3/1999 10:03:11 AM]
 Nicholson, Teri: Underwater Vocalization and Subsequent Social Beh...hardsi) Near Hopkins Marine Station, Pacific Grove, CA, MBARI 1997

estimate the true abundance of harbor seals in Washington and Oregon, 1991-1993. Abstract in Eleventh
Biennial Conference on the Biology of Marine Mammals.
Kunnasranta, M, H Hyvarinen, and J Sorionen. 1996. Underwater vocalizations of Ladoga ringed seals
(Phoca hispida ladogensis nordq.) in summertime. Mar. Mamm. Sci., 12(4):61 1-6 18.
Le Boeuf, BJ. 1991. Pinniped mating systems on land, ice and in the water: Emphasis on the Phocidae.
Behaviour of Pinnipeds (ed. D. Renouf). Chapman and Hall, London. pp. 45-65.
Lowry, LF, and KJ Frost. 1995. Movement of satellite-tagged harbor seals in Prince William Sound,
Alaska, 1991-1994. Abstract in Eleventh Biennial Conference on the Biology of Marine Mammals.
Mohl, B, JM Terhune, and K Ronald. 1975. Underwater calls of the harp seal, Pagophilus groenlandicus.
Rapp. P.-v. Reun. Cons. int. Explor. Mer., 169:533-543.
Pitcher, KW, and DC McAllister. 1981. Movements and haul out behavior of radio-tagged harbor seals,
Phoca vitulina. Canadian Field-Naturalist, 95:292-297.
Ray, GC, and WA Watkins. Social function of underwater sounds in the walrus Odobenus rosmarus.
1975 Rapp. P.-v. Reun. Cons. int. Explor. Mer., 169:524-526.
Ray, GC. WA Watkins, and JI Burns. 1969. The underwater song of Erignathus (bearded seal).
Zoologica, 54 (2):79-83.
Rogers, TL, DH Cato, and MM Bryden. 1996. Behavioral significance of underwater vocalizations of
captive leopard seals, Hydrurga leptonyx. Mar. Mamm. Sci., 12(3):414-427.
Slater, LM. and H Markowitz. 1983. Spring population trends in Phoca vitulina richardsi in two central
California coastal areas. Calif Fish and Game, 69:217-226.
Stirling, I. 1973. Vocalization in the ringed seal (Phoca hispida). J. Fish. Res. Board.Can..
30(10):1592-1594.
Stirling, I, W Calvert, and C Spencer. 1987. Evidence of stereotyped underwater vocalizations of male
Atlantic walruses (Odobenus rosmarus rosmarus). Can. J. Zool., 65:2311-2321.
Sullivan, RM. 1981. Aquatic displays and interaction in harbor seals, Phoca vitulina, with comments on
mating systems. J. Mamm., 62(4):825-831.
Thompson, PM. 1987. The effect of seasonal changes in behavior on the distribution and abundance of
common seals. Phoca vitulina, in Orkney. Unpubl. Ph.D. Dissert., Univ. of Aberdeen, England. l67 pp.
Thompson, PM. D Miller. R Cooper, and PS Hammond. 1994. Changes in the distribution and activity of
male harbor seals during the breeding season: implications for their lactation strategy and mating
patterns. J ofAnim. Ecol., 62:24-30.
Van Parijs, SM, PM Thompson, DJ Tollit, A MacKay. 1996. Changes in the distribution and activity of
male harbor seals during the mating season. Anim. Beh. in press.
Yochem, PK. BS Stewart, RL DeLong, and DP DeMaster. 1987. Diel haul out patterns and site fidelity
of harbor seals (Phoca vitulina richardsi) on San Miguel Island, California in Autumn. Mar. Mamm. Sci.,
3:323-332.

 http://www.mbari.org/interns/projects/papers/97Nicholson.html (6 of 6) [11/3/1999 10:03:11 AM]

				
DOCUMENT INFO