Content uploaded by Renata Santoro Sousa-Lima
Author content
All content in this area was uploaded by Renata Santoro Sousa-Lima on Dec 05, 2014
Content may be subject to copyright.
Aquatic Mammals
2008,
34
(1), 109-122, DOI 10.1578/AM.34.1.2008.109
Gender, Age, and Identity in the Isolation Calls of
Antillean Manatees (
Trichechus manatus manatus
)
Renata S. Sousa-Lima,
1, 4
Adriano P. Paglia,
2, 4
and Gustavo A. B. da Fonseca
3, 4, 5
1
Bioacoustics Research Program, Laboratory of Ornithology, Cornell University,
159 Sapsucker Woods Road, Ithaca, NY 14850, USA; E-mail: rs132@cornell.edu
2
Conservation International do Brasil, Avenida Getúlio Vargas, 1300, 7° Andar, 30112-021 Belo Horizonte, MG, Brasil
3
Global Environment Facility, 1818 H Street, NW, G 6-602 Washington, DC 20433, USA
4
Programa de Pós-Graduação em Ecologia, Conservação e Manejo de Vida Silvestre (PG-ECMVS),
Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais, Belo Horizonte, Brasil
5
Departamento de Zoologia, Instituto de Ciencias Biologicas, Universidade Federal de Minas Gerais,
5
Departamento de Zoologia, Instituto de Ciencias Biologicas, Universidade Federal de Minas Gerais,
5
Avenida Antonio Carlos, 6627, 31270-901 Belo Horizonte, MG, Brasil
Abstract
Empirical evidence of individual vocal recogni-
tion has been reported for the Amazonian manatee
(
Trichechus inunguis
) and the West Indian mana-
tee (
T. manatus
). Underwater vocalizations of 15
Antillean manatees (
T. m. manatus
) were recorded
to verify if this subspecies also conveys individual
information through their calls. The isolation calls
selected for analysis were digitized to measure
eight different variables. Individual vocal patterns
were analyzed within two age classes (calves and
others) and between sexes. Discriminant func-
tion analysis for each age class grouped vocal-
izations by individual, based on variables related
to the fundamental frequency and call duration.
Female calls were longer in duration and pre-
sented a higher fundamental frequency but lower
peak frequency values than males. Calves had
significantly higher values for all eight acoustic
variables measured with respect to frequency and
time. Higher values for all frequency parameters
in calf calls and the inverse relationship between
total body length and peak frequency suggests that
younger, smaller animals emit higher frequency
sounds. Furthermore, higher values obtained
for the fundamental frequency range of calves
and the inverse relationship of this variable with
total body length suggest that the fundamental
frequency becomes more defined as the animal
ages. Vocal learning and genetic inheritance are
discussed based on the analyses of vocal patterns
among related individuals. In addition to facilitat-
ing individual recognition as a possible factor in
Antillean manatee social interactions, vocal iden-
tity provides a potential means of estimating the
size and structure of sirenian populations.
Key Words:
Sirenia, manatee,
Trichechus mana-
tus manatus
, calls, sounds, vocalization, gender
differences, age differences, vocal identity,
communication
Introduction
The West Indian manatee was split into two sub-
species: (1) the Florida manatee (
Trichechus
manatus latirostris
) and (2) the Antillean mana-
tee (
T. m. manatus
) (Domning & Hayek, 1986).
This division has since been questioned by evi-
dence from mitochondrial DNA haplotypes
shared between the Florida manatee population
and populations from the Greater Antilles, which
are considered to be part of the Antillean manatee
subspecies (Garcia-Rodriguez et al., 1998). More
recently, Catanhede et al. (2005) and Vianna et al.
(2006) proposed that
T. manatus
, the West Indian
manatee, is a paraphyletic species.
These taxa and other sirenian species may pres-
ent identity information in their calls. Vocal sig-
nature information and individual recognition has
already been documented in Amazonian manatees
(
T. inunguis
) (Sousa-Lima et al., 2002). Additional
evidence stems from individual vocal differences
in dugongs (
Dugong dugon
) (Anderson & Barclay,
1995) and individual vocal recognition between a
mother and calf pair of Florida manatees that had
been physically separated (Reynolds, 1981). Florida
manatee vocalizations function to assist mothers
and calves in locating and maintaining contact with
each other (Hartman, 1979) and also may have the
potential for conveying identity information.
Animal communication signals are often accom-
panied by information about the sender such as
motivation, sex, age, or identity (Halliday & Slater,
1983). Such information presumably benefits both
parties (sender and receiver), and the acquisition
110
Sousa-Lima et al.
of these benefits is the function of the exchange
(Bradbury & Vehrencamp, 1998). Therefore, if a
signal that conveys information about gender, age,
or identity of an individual is mutually beneficial,
and no other means of recognition is reliable,
selection should act to enable individuals to dis-
criminate among such attributes. Selective forces
to identify and locate specific individuals, par-
ticularly offspring, are typically strongest in colo-
nial species because the probability of misdirect-
ing parental care increases when many offspring
are present (e.g., in bats,
Nycticeius humeralis
;
Scherrer & Wilkinson, 1993).
Despite the relatively recent large-scale reduc-
tions in numbers throughout their range (Lima,
1997), Antillean manatees were once abundant,
and groups of over 300 individuals were observed
along the Brazilian coast (Whitehead, 1978).
Florida manatees were thought to breed year-
round with a slight birth peak in spring (Hartman,
1979), but more recently Rathbun et al. (1995)
reported that Florida manatees have a more pro-
nounced calving peak between April and May.
In Brazil, the Antillean manatee calving season
occurs between October and March (Lima, 1997;
Paludo & Langguth, 2002). Manatees have low
reproductive rates and high maternal investment—
that is, long gestation, high birth weights, single
offspring (twins are rare), and extended maternal
care (O’Shea & Reep, 1990; Rathbun et al., 1995;
da Silva et al., 2000, in press). Therefore, it is rea-
sonable to assume that there is selective pressure
on mothers to nurse their own calves, rather than
other calves implying that there is an adaptive
advantage to individual recognition ability.
Recognizing and maintaining contact with spe-
cific individuals may be more challenging when the
animals are often physically separated and adequate
visual contact is not possible. The existence of indi-
vidual information and/or the capacity for individ-
ual recognition has been reported in sounds pro-
duced by other marine mammals in which mothers
and offspring are often separated and reunited such
as in pinnipeds (Insley et al., 2003). Manatees typi-
cally inhabit turbid waters (Moore, 1951). In Brazil,
Antillean manatees are observed in feeding aggre-
gations around river mouths (Paludo & Langguth,
2002), and their tendency to wander while foraging
causes unintentional separations between mothers
and calves, thus favoring the location and recogni-
tion of individual calves through sound rather than
visually (Hartman, 1979; Reynolds, 1981; Sousa-
Lima et al., 2002).
In addition to individual identity, social recog-
nition (i.e., age, gender, and reproductive status)
might be a possible function of sirenian calls.
Although mature male dugongs develop tusks, there
is no apparent sexual dimorphism in manatees. The
role of vision and olfaction in the reproduction of
fully aquatic mammals, such as sirenians, is limited
by the environment and should only be effective
at short ranges. The aquatic environment favors
the dispersal of mates in space and an increased
reliance on acoustical cues for mate location
(Anderson, 2002), recognition, and choice.
The recognition process begins with the emis-
sion of unique and stereotyped signals by the
sender. Herein we predict that the underwater
isolation calls of Antillean manatees will show
variations that can be attributed to individual,
stereotyped differences as well as gender and age
information.
Materials and Methods
Underwater sounds from eight male and six female
Antillean manatees were recorded at the facili-
ties of the Centro Mamíferos Aquáticos/IBAMA
(CMA), a governmental research center located
on Itamaracá Island in the northeastern state of
Pernambuco, Brazil. Additionally, one rehabili-
tated and reintroduced female was recorded in an
estuarine lagoon located on the southern coast of
Maceió City, state of Alagoas, Brazil. Identification
information for the animals recorded in this study
is given in Table 1. The study animals were mostly
newborn, stranded calves that had been rescued
and taken to CMA. Individuals younger than 3 y
old were considered calves based on the estimated
inter-birth interval of 3 to 5 y (da Silva et al., in
press) and on direct evidence from the Amazonian
species in which weaning takes place after 2 y of
age (da Silva et al., 2000).
There were two large pools at CMA (10.1 m in
diameter, 4.15 m in depth, and capacity for 400,000 l
of water each) connected by a smaller pool (4.45
× 4.1 × 1.2 m and capacity for 20,000 l) where
adults and a few calves were maintained. Most
calves were kept in another smaller pool (8.0 ×
5.0 × 1.5 m and capacity for 96,000 l). Two other
small pools were used for isolating calves: a round
one (3.0 × 0.8 m) and a rectangular one (4.0 × 2.0
× 0.9 m) with 7.8 and 8.0 m
3
of water capacity,
respectively.
Two age classes (calves and others) were
considered in the analysis because distinguish-
ing adults from subadults based on size is too
arbitrary (Hartman, 1979). Individual adults and
subadults were isolated in one of the main pools,
while calves were transferred on a stretcher to the
smaller pools, which were completely isolated
during recordings. Individuals were isolated for
25 min to ensure that the sample of recordings
from all individuals was taken in a similar context.
Animals could not be kept in isolation any longer
for ethical reasons, thus limiting the number of
Gender, Age, and Identity in Manatee Calls
111
suitable vocalizations acquired for analysis. A
canoe was used to transport the necessary equip-
ment and to approach the reintroduced animal
which was recorded in the wild (Lua; see Table
1) for 3 h at a distance of 5 m. This reintroduced
animal was recorded when it was by itself; there-
fore, no recordings of calls emitted in a social con-
text in the wild were obtained. Captive animals
were recorded in two different social contexts,
however: (1) in isolation (25 min per individual)
and (2) during conspecific interactions (approxi-
mately 5 h of recordings in each pool).
Recordings were made using a Sony Walkman
Pro (WM-D6C; flat audio frequency response 40 to
15,000 Hz ± 3 dB) and an omnidirectional hydro-
phone (Cetacean Research Technology Model
50Ca) with a sensitivity of -161 dB re: 1
µ
Pa and
a frequency range response of 0.01 to 310.0 kHz.
Analyzed calls were chosen based on a high signal-
to-noise ratio and on their occurrence in a similar
behavioral state (swimming). Analog recordings
were digitized at a sample rate of 44,100 Hz;
sample size 16 bits. A maximum of 10 sounds per
individual were randomly selected; some individu-
als did not vocalize enough to use a larger sample
size (see Table 1). Each sound was analyzed using
Canary 1.2.1.
software (Charif et al., 1995), with a
filter bandwidth of 699.40 Hz and a frame length
of 256 points. The duration and frequency were
measured from the most intense harmonic that
was clearly visible along the length of the signal.
We then divided the frequency measurements by
the appropriate factor (number of the harmonic) to
yield the value of the fundamental frequency.
We measured eight acoustic variables from
the recordings and six of them (listed in Table 2)
were used to perform two discriminant function
analyses (DFA), one for each age class, to test for
differences in individual vocal patterns. The vari-
ables “fundamental frequency range” and “har-
monic with most energy” were excluded from the
DFA because they did not conform to the test’s
assumptions (they were linear combinations of
other variables). In both DFAs, sounds from each
individual were treated as a group (Manly, 1994).
Two separate nested multi-way analyses of
variance (MANOVA), using all eight measured
variables (Table 3), were performed to test for dif-
ferences in the vocal pattern between sexes and
between age classes. In each analysis, the indi-
vidual vocalizations were nested by sex and age
classes. This procedure removes the effect of the
differences between individuals from the main
factors (sex and age classes).
The mean values of the variables from each
individual were incorporated into a linear regres-
sion to examine the correlation between total body
length and each of the eight vocal variables (Sokal
& Rohlf, 1995).
Results
Vocal Repertoire
Two sound types, clicks and vocalizations, were iden-
tified for this species (Figures 1 & 2). The broadband
click sounds had dominant frequencies between 1.0
to 4.0 kHz, with energy as high as 20.0 kHz. This
sound type was recorded in all contexts observed,
Table 1.
Identification of the Antillean manatees recorded in Brazil
Individual
Sex
∗
Age class
♦
Origin (location/state)
ϕ
Body length (cm)
Calls analyzed
Tibau
M
C
Tibau do Norte/RN
159
10
Xiquito
M
C
CMA – Xica’ s calf
167
10
Sheila
F
C
CMA – Sereia’ s calf
178
10
Boi Voador
M
C
São Luis/MA
182
10
Carla
F
C
CMA – Sereia’ s calf
180
10
Araqueto
M
C
Aracati/CE
190
10
Guaju
M
C
Sagi/RN
199
10
Guape
M
C
Coqueirinho/PB
208
10
Poque
M
O
Oiapoque/AP
215
10
Lua
F
O
Morro Branco/CE
240
10
Marbela
F
O
Pipa/RN
252
10
Xuxa
F
O
Morro Branco/CE
266
10
Xica
F
O
Goiana/PE
271
8
Netuno
M
O
Sagi/RN
285
10
Sereia
F
O
Barro Preto/CE
280
5
∗
M = male; F = female
♦
C = calf; O = other (subadult or adult)
ϕ
CE: Ceará; MA: Maranhão; CMA: Centro Mamíferos Aquáticos; RN: Rio Grande do Norte; PB: Paraíba; AP: Amapá
112
Sousa-Lima et al.
sometimes, but not always, in association with the
vocalizations (e.g., Figure 1: O & Figure 2).
In addition to the clicks, the vocalization rep-
ertoire was a continuum, ranging from relatively
simple sounds with a clear harmonic structure to
more complex sounds with a noisy, harsh quality
that is comprised either by a combination of har-
monic and nonharmonic (e.g., noisy, nonlinear,
chaotic) frequency components or by broad fre-
quency bands (e.g., Figure 1: M). Nonharmonic,
chaotic components occurred in the beginning, at
the end, or throughout the entire signal, although
this asynchrony in the frequency bands is mainly
observed at the beginning and end of the signal
(Figure 1: H, N, O, Q & R). All individuals emit-
ted nonharmonic and harmonic calls and, due to
the relative ease of taking measurements from the
sounds with clear harmonics, those were the ones
explored further in this study.
The vocalizations had only one component with
a mean duration of 353 ± 78 ms, ranging from 180
to 480 ms, with both amplitude and frequency mod-
ulation. The mean frequency variation (maximum-
minimum fundamental frequency) was 0.97 ± 0.50
kHz. The majority of vocalizations recorded began
with ascending frequency modulation and ended
with descending frequency modulation, although
some individuals varied in that respect and can
have different modulation features in their calls
(Figure 1: N-U). The mean fundamental frequency
of these vocalizations was 2.45 ± 0.50 kHz, with
peak frequencies ranging from 3.7 to 5.7 kHz. In
many cases, several harmonics were more intense
than the fundamental (Figure 1).
Individual Stereotypy
Each individual had a single type of clear har-
monic isolation call with some degree of variation.
Intraindividual variation is illustrated in the spectro-
grams (Figure 1) and also in the spread of the sym-
bols representing each individual in the plot of the
first two discriminant functions or axes (Figure 3).
Calves and noncalves are plotted separately in Figure
3 to enable better visualization of the data ordination
due to considerable overlap, mainly in the second
axis. The ordination of the six variables separates
individuals even though there is great intraindividual
variation and overlap between the vocal patterns of
some individuals. The first two discriminant func-
tions explained more than 93% of the common vari-
ance in the calls for both age classes. Table 2 shows
that, for both age classes, the first discriminant func-
tion is highly correlated with the mean fundamen-
tal frequency and the second highly correlated with
signal duration. Note that some overlapping individ-
uals, such as Carla and Sheila (Figure 3a), are twin
sisters. Figure 1 (N-Q & R-U) also illustrates the
similarities among the calls of these two individuals.
In the DFA with all individuals (not presented here
due to difficulty in visualization in 2D), another case
of overlap between the vocal patterns of kin arose:
between a mother and calf pair, Xiquito (Figure 3a)
and Xica (Figure 3b).
Age and Sex Discrimination
There were significant differences in the acoustic
parameters of vocalizations between sexes and age
classes (Table 3). Calves presented higher values
for all the acoustic variables except in the number
of harmonics. Females had higher values for
signal duration and mean, maximum, and mini-
mum fundamental frequencies compared to males,
but they presented fewer harmonics and a lower
mean peak frequency. No significant difference in
fundamental frequency range between males and
females was observed.
Table 2.
Factor loading of each variable on the two canonical functions of the discriminant function analysis; the variables
with highest loading are highlighted.
Calves
Others
Variable
Function 1
Function 2
Function 1
Function 2
Signal duration
0.0075
-0.8661
-0.0869
0.8966
Number of harmonics
-0.4471
-0.0805
0.3174
0.3723
Mean fundamental frequency
0.8028
-0.2034
-0.9065
-0.2567
Maximum fundamental frequency
0.4104
-0.0256
-0.4659
0.0494
Minimum fundamental frequency
0.4503
0.0914
-0.3859
-0.1586
Peak frequency
-0.0684
0.0847
0.0223
0.2856
Statistics
Cumulative probability
0.766
0.937
0.865
0.934
Wilk’s
λ
0.001
0.03
0.005
0.118
X
2
495.3
252.8
288.9
118.6
df
42
30
36
25
p
<0.001
<0.001
<0.001
<0.001
Gender, Age, and Identity in Manatee Calls
113
Figure 1.
Four examples of isolation call spectrograms from five different Antillean manatees; time on the X axis (0 to 500 ms) and frequency on the Y axis (0.0 to 20.0 kHz). “Carla”
and “Sheila” are twin sisters; note their calls’ similarity.
114
Sousa-Lima et al.
Significant inverse relationships were found
between total body length and both the fundamen-
tal frequency range (F
1, 13
= 5.49;
r
= -0.54;
p
=
0.038) and the peak frequency (F
1, 13
0.038) and the peak frequency (F
1, 13
1, 13
= 9.03;
r
=
r = r
-0.65;
p
= 0.008) (Figures 4 & 5).
Discussion
Sound Repertoire
Sound analysis and description can be a very sub-
jective exercise. For example, some authors mea-
sure slightly different things such as mean fun-
damental frequency vs the absolute value of the
fundamental frequency at some point along the
duration of the call. Thus, in order to avoid further
artifacts, we made our comparisons by using only
quantitative data provided in the literature.
Our data on the Antillean manatee vocal rep-
ertoire adds to the data gathered by Sonoda &
Takemura (1973) and Nowacek et al. (2003). The
first authors recorded two captive Antillean mana-
tees and found “click-like,” “frog-like,” and “multi-
layer” vocalizations and described clicks as very
short signals with a main frequency range from
2.0 to 7.0 kHz. We recorded clicks that reach peak
frequencies up to 14.0 kHz (Figure 2). Nowacek
et al. (2003) recorded several individuals and did not
report clicks in wild manatees. Clicks and vocaliza-
tions have also been recorded in semi-captivity in
Brazil (approximately 250 h of recordings by Sousa-
Lima, unpub. data). There is no information about
how the click sounds described for this species are
produced. Although speculative, we believe that the
click sounds may be generated by teeth movements
rather than by a specific anatomical structure as in
cetaceans (Cranford et al., 1996).
Sonoda & Takemura’s (1973) “frog-like” and
“multi-layer” vocalizations are similar to the calls
recorded by Nowacek et al. (2003) and those in
Figure 1. Sonoda & Takemura (1973) indicated
durations of less than 200 ms and a main frequency
range of 2.0 to 4.0 kHz, while a more recent paper
describing these vocalizations (Nowacek et al.,
2003) indicates a wider range for the duration (32
to 217 ms). The shortest call recorded in Brazil
was 110 ms (Sousa-Lima, 1999), and the duration
limits we found in this study were 180 to 480 ms.
To summarize, taking account of all the informa-
tion available, Antillean manatee vocalizations
have durations as short as 32 ms (Nowacek et al.,
2003) and as long as 480 ms (this study).
The range of the fundamental frequency
reported here (1.07 to 4.98 kHz) is consistent
with the data available in the literature (Sonoda
& Takemura, 1973; Nowacek et al., 2003). The
mean peak frequency range (3.7 to 5.7 kHz) was
also consistent with the range (3.18 to 7.08 kHz)
reported by Nowacek et al. (2003). To our knowl-
edge, no other data have been published on the
repertoire of Antillean manatees, and there are no
obvious differences in the vocalizations produced
by manatees from Florida and Belize (Nowacek
et al., 2003). Additionally,
T. manatus
is consid-
ered a paraphyletic species (Catanhede et al., 2005;
Vianna et al., 2006); therefore, we also examined
the data available for manatees from Florida.
Steel (1982) reported duration limits of 80 to
470 ms for Florida manatees, while Schevill &
Watkins (1965) found limits of 150 to 500 ms.
Fundamental frequency values are very similar
between our study and that of Steel (1.08 to 5.00
kHz). Schevill & Watkins reported a lower limit
for the fundamental frequency (0.6 kHz). Peak fre-
quencies range from 1.0 to 12.0 kHz (Steel, 1982).
No click-like sounds have been reported for the
Florida manatee. Despite differences in measure-
ment methods and populations, we can reasonably
conclude that West Indian manatees have vocal-
izations that range from 32 to 500 ms in duration,
with a fundamental frequency of 0.6 to 5.0 kHz and
peak frequencies ranging from 1.0 to 12.0 kHz.
It is not yet clear how the nonharmonic fre-
quency bands in the vocalizations are produced
and, in some cases, there is a suggestion of two
different sources of sound production (Figure
1: N, R & S) (also noted by Mann et al., 2006).
Manatees (both the Antillean and the Amazonian
species) contract the muscles dorsal to the max-
illa and caudal to the nostrils while maintaining
the nostrils closed when vocalizing (Sousa-Lima,
pers. obs.). It is possible that the contraction of
those muscles (probably the naso-labial or the
maxillo-labial muscles described for West African
manatees [
T. senegalensis
] by Saban, 1975) might
Figure 2.
Example of clicks produced by Antillean
manatees
Gender, Age, and Identity in Manatee Calls
115
contribute to the acoustic characteristics of the
sounds produced.
Sex Differences in Acoustic Behavior and Mating
Strategies
We found significant differences between sexes in
Figure 3.
Plot of the first two DFA axes: Discriminant Axis 1 is related to frequency and number of harmonics; Discriminant
Axis 2 is related to signal duration and peak frequency (function of the relative intensity contribution of the harmonics).
Calves (a) and Others (b) were plotted separately to enable better visualization of individual groupings.
116
Sousa-Lima et al.
Table 3.
Results of the nested MANOVA, mean ± SD, and multivariate F-test between age classes and sexes of each variable;
all variables, except that “signal duration” was log-transformed for the statistical analysis. The mean values presented are in
the original linear scale.
Age class
Sex
(Wilk’s
λ
= 0.103; F
7, 122
= 152.2;
p
< 0.001)
(Wilk’s
λ
= 0.290; F
7, 122
= 42.6;
p
< 0.001)
Variable
Calves
(
n
= 80)
Others
(
n
= 63)
F
1; 128
p
-level
Males
(
n
= 80)
Females
(
n
= 63)
F
1; 132
p
-level
Signal duration (ms)
371.9 ±
80.4
323.9 ±
78.6
25.0
< 0.001
320.7 ±
82.4
388.8 ±
66.5
56.3
< 0.001
Number of harmonics
6.2 ± 1.5
7.1 ± 1.4
50.6
< 0.001
6.9± 1.4
6.2 ± 1.6
36.1
< 0.001
Harmonic with most energy
2.2 ± 1.4
1.8 ± 1.2
6.1
0.015
2.3 ± 1.2
1.7 ± 1.2
18.8
< 0.001
Mean fundamental
frequency (kHz)
2.7 ± 0.5
2.1 ± 0.4
1,042.0
< 0.001
2.4 ± 0.5
2.6 ± 0.6
114.7
< 0.001
Maximum fundamental
frequency (kHz)
3.3 ± 0.7
2.5 ± 0.6
231.7
< 0.001
2.8 ± 0.7
3.1 ± 0.8
39.4
< 0.001
Minimum fundamental
frequency (kHz)
2.2 ± 0.5
1.8 ± 0.4
109.8
< 0.001
1.9 ± 0.5
2.1 ± 0.5
50.3
< 0.001
Fundamental frequency
range (kHz)
1.2 ± 0.5
0.7 ± 0.3
60.2
< 0.001
0.9 ± 0.4
1.0 ± 0.5
0.4
0.530
Peak frequency (kHz)
5.7 ± 2.9
3.7 ± 2.2
26.6
< 0.001
5.4 ± 3.1
4.1 ± 2.2
11.1
0.001
Figure 4.
Regression between total body length and peak frequency of calls made by Antillean manatees
Gender, Age, and Identity in Manatee Calls
117
Antillean manatee calls. Females emit calls with
higher fundamental frequencies but lower peak
frequencies than males. Also female calls are
longer in duration than male calls. Steel (1982)
reports higher pitched calls (higher fundamental
and peak frequencies) for male Florida manatees.
However, Steel was unable to confirm the identi-
fication or location of the vocalizing individual,
which could have resulted in misidentification
of the animal producing the sound. Nonetheless,
sexual vocal differences suggest that the emission
of calls may have a function in reproduction.
Anderson (2002) lists gaps in information about
the manatee’s sexual strategies and indicates that
one of these is “how males recognize estrous,”
suggesting that estrous females may advertise
their receptiveness acoustically. Bengston (1981)
showed that adult male Florida manatees estab-
lish regular search circuits in areas frequently
used by females in a supposed attempt to assess
female reproductive status. We suggest that fre-
quent calling between mother and calf pairs
might aid roving males in locating females. This
presumption is based on observations by Hartman
(1979) and O’Shea & Hartley (1995) that Florida
manatee females with young calves are harassed
by “mating” herds of males and that that could be
considered as a form of infanticide to induce sub-
sequent estrous.
Morton’s (1977) motivation-structural rules
suggest that increased emission rates and higher
frequency calls are both characteristics of signals
that elicit the approach of receivers. We can spec-
ulate that females could be using their high fre-
quency calls to attract males to their location. In
the Antillean as well as in the Amazonian manatee
(Sousa-Lima et al., 2002), females present greater
fundamental frequency values than males (i.e.,
they tend to emit higher frequency vocalizations
despite their bigger size). Furthermore, female
Antillean manatees emit longer isolation calls
than males. Although we do not know if the indi-
viduals recorded were in estrous, we speculate that
females could also be altering other parameters in
their calls (such as signal duration or intensity)
to inform males of their receptiveness. Female
estrous vocal cues, such as increased call inten-
sity, have been reported in elephants (
Loxodonta
africana
) (Poole et al., 1988).
Figure 5.
Regression between total body length and range of fundamental frequency of calls made by Antillean manatees
118
Sousa-Lima et al.
Age Differences
Signaling strategies such as increased emission
rates and higher frequency calls are also widely
used in parent-offspring interactions in mammals
and birds (Morton, 1977). In isolation, Amazonian
manatee calves vocalize more than adults: 6 calls/
min vs l call/min (or not at all) during a 25-min
recording session (Sousa-Lima et al., 2002). We
found no difference in calling rates between age
classes or sexes in Antillean manatees recorded
in isolation (3 calls/min); however, context and
stress levels may influence call emission rate in
Antillean manatees. Individuals recorded by us in
isolation present a higher calling rate than individ-
uals in the wild recorded by Nowacek et al. (2003)
(0.09 to 0.75 calls/min). Individual calling rates
in the wild Florida manatee are also lower, rang-
ing between 0.25 to 4.75 calls/5 min (Bengston &
Fitzgerald, 1985; Phillips et al., 2004; Miksis-Olds,
2006). Furthermore, other environmental charac-
teristics, such as elevated noise levels, increase
the Florida manatee’s calling rate (Miksis-Olds,
2006). Manatee calves might have different spe-
cies-specific strategies to call conspecific atten-
tion (mainly their mothers’). Amazonian manatee
calves increase the number of vocalizations emit-
ted (Sousa-Lima et al., 2002), while the calves
in the Antillean species emit longer and higher
pitched calls (this study).
The inverse correlation between the funda-
mental frequency range of the calls and total
body length in Antillean manatees indicates that
maturity has an effect of decreasing this vocal
parameter. As noted elsewhere (Sousa-Lima et al.,
2002), it appears that as the manatee grows older,
the fundamental frequency range decreases, thus,
the most important parameter that confers vocal
identity (fundamental frequency) is better defined
as the animal ages. This phenomenon is observed
in adolescent male humans and in birds such as
penguins (
Aptenodytes patagonicus
penguins (Aptenodytes patagonicuspenguins (
) as they pre-
pare to leave the nest (Robisson, 1992).
Interindividual Discrimination
Some of the most important variables for individ-
ual vocal discrimination in Antillean manatees are
related to the fundamental frequency as also has
been found for pinnipeds (Insley, 1992; Charrier
et al., 2003; Insley et al., 2003). Nevertheless,
DFA function 2 (related to signal duration; Figure
3) also contributed to the isolation of individual
vocal patterns. Signature information in Antillean
manatees can also be contained in the frequency
modulation of the calls (not analysed in detail
here). Nevertheless, vocal identity involves
the more easily measured features of duration;
number of harmonics; and minimum, maximum,
and mean fundamental frequency.
Other characteristics of the signal may carry
additional identity information but could not
be included in the quantitative analysis because
they did not conform to the test assumptions or
were not measured. One of these characteristics
is the energy distribution across frequency bands,
which gives a distinct timbre or tonal quality to
the call (Marler, 1969). This unaccounted vocal
feature can be observed in Figure 1 and may also
be perceived as an identity cue by the manatees.
Therefore, individual differences can be present
in both quantitative and qualitative characteristics
of the signal. At this point, we cannot say which
characteristics the animals use in the process of
discrimination/identification, but probably the
signal is perceived as a whole and differences in
qualitative and quantitative features are taken into
account during recognition.
Holekamp et al. (1999) demonstrated that
female hyenas (
Crocuta crocuta
) identify cubs as
their own vs others based on long distance vocal-
izations; whereas vervet monkeys (
Cercopithecus
aethiops
) can identify third-party relationships,
showing an impressive level of social cognition
(Seyfarth & Cheney, 1994). Although manatees
are also long-lived mammals that bear few off-
spring, which require a long period of dependence
on the mother, they are subject to less selective
pressure to recognize conspecifics within their less
social system in such a way that vocal discrimina-
tion may not be so extreme. If similar benefits can
be derived by employing a simpler coding scheme
with lower costs, one would expect selection to
favour the simpler scheme (i.e., binary recognition
task) (Bradbury & Vehrencamp, 1998). Indirect
evidence of the binary recognition hypothesis for
mother-calf manatee pairs comes from the vocal
similarity observed between the twins, Carla and
Sheila (Figure 1). The mother (Sereia) would nurse
either individual by recognizing only one vocal pat-
tern as “own calf.” The question of whether or not
Antillean manatees may be able to recognize the
voices of multiple individuals within their species
is beyond the scope of this paper. Interindividual
discrimination can be further considered in two
different ways: (1) the level of stereotypy within
an individual’s call and (2) the level of discrimina-
tion among the calls of different individuals (i.e.,
how similar the calls are from a particular individ-
ual and how different they are in comparison to the
calls of conspecifics).
If selective pressures result in a greater stereo-
typy in the individual calls of a species, we cannot
expect that manatee isolation calls would be as
readily discriminated as bat pup calls (100% of
calls classified correctly in Scherrer & Wilkinson,
1993). Bats are subjected to a much greater
selective pressure because of their spatially
Gender, Age, and Identity in Manatee Calls
119
confined nursery colony with hundreds of pups
in close proximity. In an aquatic environment,
animals are consistently more isolated from each
other so that there is ultimately less selection on
vocal patterns. Inasmuch, the degree of selective
pressure within aquatic species may also vary,
favouring more or less individual call stereotypy.
Different levels of call stereotypy were found in
fur seal species of the genus
Arctocephalus
as a
result of differential selective pressure on mother-
pup recognition (Page et al., 2002).
Similarities and Differences Between Kin
Related individuals had similar acoustic param-
eters in their vocalizations. Carla and Sheila have
very close values of DFA factors 1 and 2 (Figure
3). Note also the similar spectrogram contours
and fundamental frequency values of these two
individuals (Figure 1). Xica and Xiquito also have
similar values of both DFA axes (Figure 3). The
first two individuals are twin sisters and the other
two are a mother and calf pair. There seems to
be a strong genetic component in the vocal pat-
tern expressed by an individual manatee, which
also has been suggested for the Amazonian spe-
cies based on one case of mother-calf similarity
(Sousa-Lima et al., 2002). Nevertheless, there
were no similarities between the twins’ vocal pat-
terns and their mother’s (Sereia; Figure 3).
Bottlenose dolphin (
Tursiops truncatus
) infants
develop their own signature whistle within their
first few months of life (Caldwell & Caldwell,
1979), and there is accumulating evidence that
the development of signature whistles is strongly
influenced by vocal learning (Janik, 1999). Captive
bottlenose dolphin calves often develop whistles
that are similar to the whistles used by the human
trainer or similar to whistles of their pool mates
but not to those of their mothers (Tyack, 1997).
If manatees are capable of vocal learning, mana-
tee calves in captivity that are exposed to vocal
templates of unrelated individuals might develop
a vocal pattern different from the primarily inher-
ited one. Carla and Sheila were raised in a com-
munal pen with several other individuals, while
Xiquito was rejected by its mother and bottle-
fed in an isolated pen. It is possible that because
Xiquito was isolated from other individuals, he
retained his mother’s vocal pattern as his template
(also observed in Amazonian manatees by Sousa-
Lima et al., 2002). The influence of vocal learn-
ing might be a likely explanation; however, vocal
similarity between mothers and their male calves
may also be the result of an inbreeding avoidance
mechanism as suggested for bottlenose dolphins
(Sayigh et al., 1995) and fur seals (
Arctocephalus
(Sayigh et al., 1995) and fur seals (Arctocephalus (Sayigh et al., 1995) and fur seals (
australis
) (Phillips, 1998).
Intraindividual Variation
Individual Antillean manatees have their own
characteristic vocalization patterns, with minor
variations within an individual’s vocal signa-
ture. The intraindividual vocalization repertoire
in Antillean manatees is formed by a continuum
of signals that vary slightly and that may carry
different or additional information. Anderson &
Barclay (1995) identified several types of vocal-
izations produced by another sirenian species—
the dugong. These authors also showed that the
dugong call repertoire is a continuum rather than
a set of discrete types, with gradation between one
general type of sound to the other. Noisier sounds
have been attributed to Florida manatee calves
(Steel, 1982). These results contrast with our find-
ings for Antillean manatees, which, regardless of
age class, may emit noisy or clear sounds as part
of their repertoire.
Conservation Implications
Studies of captive animals in parallel with efforts
in the wild can provide measures for use in the con-
servation of the sirenians. Our results could form
the basis for the development of alternative tech-
niques for management and conservation of the
Antillean manatee in Brazil. Acoustic surveys of
wild populations may provide an additional means
by which to assess the population’s status along
the coast, estimated to be less than 300 individuals
(Lima, 1997). The application of vocal identity in
assessing population estimates and structure in the
wild should be explored as a potential tool in the
conservation of this species, which is considered
the most endangered mammal in Brazil (listed as
“critically endangered”; IBAMA, 2001).
An improved understanding of the natural
variation found in manatee calls might be useful
in developing acoustic detectors to warn boaters
of the presence of manatees. The development of
passive acoustic detectors of manatee calls (see
algorithms tested in Niezrecki et al., 2003) and
other improvements on this technology (Yan et al.,
2005) would decrease the number of boat strikes,
which cause the injury or death of these highly
threatened animals.
In addition to more accurately estimating
population size and structure, and helping avoid
boat strikes, another pressing conservation issue
along the Brazilian coast is the increasing rate of
stranded calves (IBAMA, 2001). Lima (1997) sug-
gested that females are no longer able to give birth
inside sediment-filled estuaries due to erosion;
therefore, calving must occur in the sea where
breaking waves might make it more difficult for
calves to maintain close contact with theirs moth-
ers and which might mask their isolation calls due
120
Sousa-Lima et al.
to greater ambient noise levels, preventing their
location and reunion after separation.
Mothers may stay in the vicinity of the area where
the lost calf was stranded for periods of over 12 h
(Sousa-Lima, pers. obs.). Playbacks of vocaliza-
tions of stranded calves that have been rescued may
help attract their mothers and allow the release of
the calf
in loco
the calf in locothe calf
, significantly decreasing the costs of
rehabilitation in captivity and promoting the main-
tenance of the wild population without complicated
and costly reintroduction efforts. Manatees may be
“ecologically trapped”—that is, they may present
poor habitat choice based on cues that formerly cor-
related to habitat quality (Schlaepfer et al., 2002).
Manatee mothers choosing to give birth inside
calmer waters of estuarine rivers might be experi-
encing reduced reproductive success in an environ-
ment that has been altered by humans. The adaptive
value of the evolution of isolation calls as means of
identification, location, and proximity maintenance
between mothers and calves may have decreased as
a consequence of habitat destruction.
Acknowledgments
We thank the CMA for the opportunity to work
with the animals and for their support during this
project, especially Régis P. Lima, Jocyere Vergara,
Denise F. Castro, Danielle Paludo, and Francisco
A. P. Colares. We thank Cícero de Oliveira, Ricardo
Capetinga, and Simone M. Miranda for field assis-
tance. Comments from Jason A. Mobley, Anthony
B. Rylands, Christopher W. Clark, Jack Bradbury,
the reviewers, and the editor greatly improved
this work. This research was funded by Fundação
O Boticário de Proteção à Natureza/MacArthur
Foundation and Conservation International do
Brasil. U.S. Fish & Wildlife Service partially
funded the graduate program where RSS-L
received her Master’s degree, with a scholarship
from the Brazilian Council for Scientific and
Technological Development (CNPq).
Literature Cited
Anderson, P. K. (2002). Habitat, niche, and evolution
of sirenian mating systems.
Journal of Mammalian
Evolution
,
9
, 55-98.
Anderson, P. K., & Barclay, R. M. R. (1995). Acoustic
signals of solitary dugongs: Physical characteristics
and behavioral correlates.
Journal of Mammalogy
,
76
,
1226-1237.
Bengston, J. L. (1981).
Ecology of manatees (
Trichechus
manatus
) in the St. Johns River, Florida.
Ph.D. disserta-
tion, University of Minnesota, St. Paul.
Bengston, J. L., & Fitzgerald, S. M. (1985). Potential role
of vocalizations in West Indian manatees.
Journal of
Mammalogy
,
66
(4), 816-819.
Bradbury, J. W., & Vehrencamp, S. L. (1998).
Principles of
animal communication.
Sunderland, MA: Sinauer.
Caldwell, M. C., & Caldwell, D. K. (1979). The whistle
of the Atlantic bottlenosed dolphin (
Tursiops truncatus
):
Ontogeny. In H. E. Winn & B. L. Olla (Eds.),
Behavior
of marine animals: Current perspectives in research
(Vol. 3, pp. 369-401). New York: Plenum Press.
Catanhede, A. M., da Silva, V. M. F., Farias, I. P., Hrbek,
T., Lazzarini, S. M., & Alves-Gomes, J. (2005).
Phylogeography and population genetics of the endan-
gered Amazonian manatee,
Trichechus inunguis
,
Natterer, 1883 (Mammalia, Sirenia).
Molecular Ecology
,
14
, 401-413.
Charif, R. A., Mitchell, S., & Clark, C. W. (1995).
Canary
1.2.1 user’s manual.
Ithaca, NY: Cornell Lab of
Ornithology.
Charrier, I., Mathevon, N., & Jouventin, P. (2003).
Individuality in the voice of fur seal females: An anal-
ysis study of the pup attraction call in
Arctocephalus
tropicalis
.
Marine Mammal Science
,
19
, 161-172.
Cranford, T. W., Amundin, M., & Norris, K. S. (1996).
Functional morphology and homology in the odonto-
cete nasal complex: Implications for sound generation.
Journal of Morphology
,
228
, 223-285.
da Silva, V. M. F., Sousa-Lima, R. S., & Luna, F. (In press).
Trichechus manatus
Linnaeus, 1758. In A. Machado,
G. Drummond, & C. S. Martins (Eds.),
Livro ver-
melho da fauna Brasileira ameaçada de extinção.
Belo
Horizonte, MG, Brazil: Fundação Biodiversitas.
da Silva, V. M. F., D’Affonseca Neto, A., Mattos, G. E.,
& Sousa-Lima, R. S. (2000). Duração da lactação em
peixe-boi da Amazônia (
Trichechus inunguis
): Estudo
de caso de filhote nascido em cativeiro.
Resumos IX
Reunión de Trabajo de Especialistas em Mamíferos
Acuáticos de América del Sur.
Buenos Aires, Argentina.
Domning, D. P., & Hayek, L. C. (1986). Interspecific
and intraspecific morphological variation in manatees
(Sirenia:
Trichechus
).
Marine Mammal Science
,
2
,
87-144.
Garcia-Rodriguez, A. I., Bowen, B. W., Domning, D. P.,
Mignucci-Giannoni, A. A., Marmontel, M., Montoya-
Ospina, R. A., et al. (1998). Phylogeography of the
West Indian manatee (
Trichechus manatus
): How many
populations and how many taxa?
Molecular Ecology
,
7
,
1137-1149.
Halliday, T. R., & Slater, P. J. B. (1983). Introduction. In
T. R. Halliday & P. J. B. Slater (Eds.),
Animal behav-
iour: Communication
(Vol. 2, pp. 1-8). Oxford, UK:
Blackwell Scientific Publications.
Hartman, D. S. (1979).
Ecology and behavior of the mana-
tee,
Trichechus manatus
in Florida
(Special Publication
5)
.
New York: American Society of Mammalogists.
Holekamp, K. E., Boydston, E. E., Szykman, M., Graham,
I., Nutt, K. J., Birch, S., et al. (1999). Vocal recognition in
the spotted hyena and its possible implications regarding
the evolution of intelligence.
Animal Behaviour
,
58
,
383-395.
Gender, Age, and Identity in Manatee Calls
121
Insley, S. J. (1992). Mother-offspring separation and acous-
tic stereotypy: A comparison of call morphology in two
species of pinnipeds.
Behaviour
,
120
, 103-122.
Insley, S. J., Phillips, A. V., & Charrier, I. (2003). A review
of social recognition in pinnipeds.
Aquatic Mammals
,
29
(2), 181-201.
IBAMA. (2001).
Mamiferos aquaticos do Brasil: Plano de
acao, II.
Brasilia, Brazil: Edicoes IBAMA.
Janik, V. M. (1999). Origins and implications of vocal learn-
ing in bottlenose dolphins. In H. O. Box & K. R. Gibson
(Eds.),
Mammalian social learning: Comparative and
ecological perspectives
(pp. 308-326). Cambridge, UK:
Cambridge University Press.
Lima, R. P. (1997).
Peixe-boi marinho (
Trichechus mana-
tus
): Distribuição, status de conservação e aspectos
tradicionais ao longo do litoral do nordeste do Brasil.
Master’s thesis, Universidade Federal de Pernambuco
Recife, Brazil.
Manly, B. J. F. (1994).
Multivariate statistical methods: A
primer.
London: Chapman and Hall.
Mann, D. A., O’Shea, T. J., & Nowacek, D. P. (2006).
Nonlinear dynamics in manatee vocalizations.
Marine
Mammal Science
,
22
(3), 548-555.
Marler, P. (1969). Tonal quality of bird sounds. In R. A.
Hinde (Ed.),
Bird vocalizations: Their relation to cur-
rent problems in biology and psychology
(pp. 5-18).
Cambridge, UK: Cambridge University Press.
Miksis-Olds, J. L. (2006).
Manatee response to
environmental noise.
Ph.D. dissertation, University of
Rhode Island, Providence. 245 pp.
Moore, J. C. (1951). The status of the manatee in the
Everglades National Park, with notes on its natural his-
tory.
Journal of Mammalogy
,
32
, 22-36.
Morton, E. (1977). On the occurrence and significance of
motivation-structural rules in some bird and mammal
sounds.
American Naturalist
,
111
, 855-869.
Niezrecki, C., Phillips, R., Meyer, M., & Beusse, D. O.
(2003). Acoustic detection of manatee vocalizations.
Journal of the Acoustical Society of America
,
114
,
1640-1647.
Nowacek, D. P., Casper, B. M., Wells, R. S., Nowacek,
S. M., & Mann, D. A. (2003). Intraspecific and geo-
graphic variation of West Indian manatee (
Trichechus
manatus
spp.) vocalizations.
Journal of the Acoustical
Society of America
,
114
, 66-69.
O’Shea, T. J., & Hartley, W. C. (1995). Reproduction and
early-age survival of manatees at Blue Spring, upper St.
Johns River, Florida. In T. J. O’Shea, B. B. Ackerman, &
H. F. Percival (Eds.),
Population biology of the Florida
manatee
(pp. 157-170). Gainesville, FL: National
Biological Service Information and Technology Report
1.
O’Shea, T. J., & Reep, R. L. (1990). Encephalization quo-
tients and life-history traits in the Sirenia.
Journal of
Mammalogy
,
71
, 534-543.
Page, B., Goldsworthy, S. D., & Hindell, M. A. (2002).
Individual vocal traits of mother and pup fur seals.
Bioacoustics
,
13
, 121-143.
Paludo, D., & Langguth, A. (2002). Use of space and
temporal distribution of
Trichechus manatus manatus
Linnaeus in the region of Sagi, Rio Grande do Norte
State, Brazil (Sirenia, Trichechidae).
Revista Brasileira
de Zoologia
,
19
, 205-215.
Phillips, A. V. (1998).
Vocal communication and mother-
pup interactions in the South American fur seal,
Arctocephalus australis. Master’s thesis, University of
Alberta, Edmonton, AB, Canada. 215 pp.
Phillips, R., Niezercki, C., & Beusse, D. O. (2004).
Determination of West Indian manatee vocalization
levels and rate.
Journal of the Acoustical Society of
America
,
115
(1), 422-428.
Poole, J. H., Payne, K., Langbauer, W. R., Jr., & Moss, C.
(1988). The social contexts of some very low frequency
calls in African elephants.
Behavioral Ecology and
Sociobiology
,
22
, 385-392.
Rathbun, G. B., Reid, J. P., Bonde, R. K., & Powell, J. A.
(1995). Reproduction in free-ranging Florida mana-
tees. In T. J. O’Shea, B. B. Ackerman, & H. F. Percival
(Eds.),
Population biology of the Florida manatee
(pp.
135-156). Gainesville, FL: National Biological Service
Information and Technology Report 1.
Reynolds, J. E., III. (1981). Aspects of the social behav-
iour and herd structure of a semi-isolated colony of West
Indian manatees
Trichechus manatus
.
Mammalia
,
45
,
431-451.
Robisson, P. (1992). Roles of pitch and duration in the dis-
crimination of the mate’s call in the penguin
Aptenodytes
patagonicus
.
Bioacoustics
,
4
, 25-36.
Saban, R. (1975). La musculature peaucière de la tête
chez un jeune Lamatin du Sénégal (
Trichechus sen-
egalensis
Link 1795; Mammifère, Sirénien).
Zentralbl
Veterinarmed
,
4
(3), 232-248.
Sayigh, L. S., Tyack, P. L., Wells, R. S., Scott, M. D., &
Irvine, A. B. (1995). Sex difference in signature whistle
production of free-ranging bottlenose dolphins,
Tursiops
truncatus
.
Behavioral Ecology and Sociobiology
,
36
,
171-177.
Scherrer, J. A., & Wilkinson, G. S. (1993). Evening bat
isolation calls provide evidence for heritable signatures.
Animal Behaviour
,
46
, 847-860.
Schevill, W. E., & Watkins, W. A. (1965). Underwater calls
of
Trichechus
(manatee).
Nature
,
205
, 373-374.
Schlaepfer, M. A., Runge, M. C., & Sherman, P. W. (2002).
Ecological and evolutionary traps.
Trends in Ecology
and Evolution
,
17
, 474-480.
Seyfarth, R. M., & Cheney, D. L. (1994). The evolution of
social cognition in primates. In L. Real (Ed.),
Behavioral
mechanisms in evolutionary ecology
(pp. 371-389).
Chicago: University of Chicago Press.
Sokal, R. R., & Rohlf, F. J. (1995).
Biometry
. New York:
W. H. Freeman and Company.
Sonoda, S., & Takemura, A. (1973). Underwater sounds
of the manatees,
Trichechus manatus manatus
and
Trichechus inunguis
(Trichechidae).
Reports of the
Institute for Breeding Research, Tokyo: University of
Agriculture
,
4
, 19-24.
122
Sousa-Lima et al.
Sousa-Lima, R. S. (1999).
Comunicacão acústica em
peixes-boi (Sirenia: Trichechidae): Repertório, discrim-
inacão vocal e aplicacões no manejo e conservacãodas
espécies no Brasil.
Master’s thesis, Universidade Federal
de Minas Gerais, Belo Horizonte, MG, Brasil. +75 pp.
Sousa-Lima, R. S., Paglia, A. P., & da Fonseca, G. A. B.
(2002). Signature information and individual recog-
nition in the isolation calls of Amazonian manatees,
Trichechus inunguis
(Mammalia: Sirenia).
Animal
Behaviour
,
63
, 301-310.
Steel, C. (1982).
Vocalization patterns and corresponding
behavior of the West Indian manatee
(
Trichechus mana-
tus
).
PhD dissertation, Florida Institute of Technology,
Melbourne.
Tyack, P. L. (1997). Development and social functions of
signature whistles in bottlenose dolphins
Tursiops trun-
catus
.
Bioacoustics
,
8
, 21-46.
Vianna, J. A., Bonde, R. K., Caballero, S., Giraldo, J. P.,
Lima, R. P., Clark, A., et al. (2006). Phylogeography,
phylogeny and hybridization in trichechid sirenians:
Implications for manatee conservation.
Molecular
Ecology
,
15
, 433-447.
Whitehead, P. J. P. (1978). Registros antigos da presença
do Peixe-Boi do Caribe (
Trichechus manatus
) no Brasil.
Acta Amazonica
,
8
, 497-506.
Yan, Z., Niezrecki, C., & Beusse, D. O. (2005). Background
noise cancellation for improved acoustic detection of
manatee vocalizations.
Journal of the Acoustical Society
of America
,
117
(6), 3566-3573.