Evidence for signature whistles in Guiana dolphins (Sotalia guianensis) in Ilheus, northeastern Brazil

Transcription

1 Evidence for signature whistles in Guiana dolphins (Sotalia guianensis) in Ilheus, northeastern Brazil Alice Lima and Yvonnick Le Pendu a) Universidade Estadual de Santa Cruz, Departamento de Ci^encias Biologicas, Rodovia Jorge Amado, km 16 Bairro Salobrinho, Ilheus, Bahia , Brazil (Received 14 July 2014; revised 7 October 2014; accepted 20 October 2014) Signature whistles have been widely studied in bottlenose dolphins (Tursiops truncatus). A recent study suggested the occurrence of signature whistles in Guiana dolphins (Sotalia guianensis) but could not identify the whistlers. The objective of this study is to describe the whistle characteristics in the population of S. guianensis from Ilheus and investigate the occurrence of signature whistles. Dolphins from 55 groups were photographed and sound emissions from 21 groups were recorded. The frequency parameters and duration of the 847 recorded whistles were similar to those recorded in 12 other populations, on an intermediate position of a latitudinal gradient. The visual classification method was applied to the spectrograms of 68 stereotyped potential signature whistles. Five out of 6 human judges agreed on the formation of 13 groups. The presence of the same individuals in different recording occasions of stereotyped whistles suggests that some whistle types are produced by specific individuals. The study is the first to use the photo-identification technique to identify Guiana dolphins emitting whistles and the results reinforce the hypothesis of signature whistles in this species. VC 2014 Acoustical Society of America.[ PACS number(s): Gf, Ka, Lb [WWA] Pages: I. INTRODUCTION To ensure the benefits of group living, like improved access to food and decreased risk of predation (Connor, 2000), animals develop mechanisms to maintain group cohesion. These mechanisms include visual signs, acoustic exchange of information, and recognition of particular individuals (Antunes et al., 2011). In dolphins, tonal sounds are known as whistles and are emitted in social interactions that involve group cohesion, individual recognition, and group feeding (May-Collado and Wartzok, 2008). Since the evolutionary history of complex tonal sounds is intertwined with social communication, May-Collado et al. (2007) suggested that complex tonal sounds may provide individual information that can facilitate the dynamics of complex societies. Caldwell and Caldwell (1965) observed that isolated captive bottlenose dolphins produced whistles with distinct frequency contours. They hypothesized that whistles are used to broadcast individual identity and named them signature whistles. According to these authors, each animal tends to emit a unique and distinct whistle irrespective of the circumstances. Research on signature whistles has been carried out with captive (Caldwell and Caldwell, 1965, 1968) and free-ranging common bottlenose dolphins (T. truncatus) (e.g., Buckstaff, 2004; Cook et al., 2004; Janik, 2000; Janik et al., 2013; King et al., 2013). Signature whistles have also been described in T. aduncus (Gridley et al., 2014), Lagenorhynchus obliquidens (Caldwell and Caldwell, 1970), Stenella frontalis (Caldwell and Caldwell, 1970; Caldwell et al., 1973; Herzing 1996), Sousa chinensis (Parijs a) Author to whom correspondence should be addressed. Electronic mail: yvonnick@uesc.br and Corkeron, 2001), and Monodon monoceros (Shapiro, 2006). Whistles are the most studied type of sound in S. guianensis (Erber and Sim~ao, 2004; Azevedo and Van Sluys, 2005; Pivari and Rosso, 2005; Rossi-Santos and Podos, 2006; Figueiredo and Sim~ao, 2009; Deconto and Monteiro- Filho, 2013). However, most of these studies have focused on the characterization of whistles while their function remains poorly understood. Whistles show great variation among populations, a feature explained in terms of dispersal limitations between populations (Rossi-Santos and Podos, 2006; May-Collado and Wartzok, 2009) or environmental characteristics of the habitats in which the population occurs (Deconto and Monteiro-Filho, 2013). Whistles frequencies and duration also vary according to behavioral state (May- Collado, 2013). May-Collado and Wartzok (2009) found that the Costa Rican Guiana dolphin population emits whistles with higher fundamental frequency and harmonics than their conspecifics in Brazil (up to and khz, respectively). Figueiredo and Sim~ao (2009) investigated the possible occurrence of signature whistles in the population of Guiana dolphins in Sepetiba Bay, Rio de Janeiro, using stereotypic contours and loop sequences as evidences of signature whistle characteristics. They identified 27 whistle types, 17 of which were recorded in a single day. The interval between successive emissions of 1 of these 17 whistles was short, suggesting that each was produced by a single animal. Most of the potential signature whistlers were traveling or foraging, suggesting a contact maintenance function of these whistles. Nonetheless, the authors did not validate their methodology to identify signature whistles: The visual inspection of the whistles was not conducted by independent human judges, nor were the identities of the whistlers 3178 J. Acoust. Soc. Am. 136 (6), December /2014/136(6)/3178/8/$30.00 VC 2014 Acoustical Society of America

2 established; thus, this study was not able to confirm whether the analyzed sounds were signature whistles. Investigating the signature whistles of free-ranging animals is difficult and most of the studies involve captive animals or capture-release events (Miksis et al., 2002). The application of the photo-identification technique to small groups of free-ranging dolphins to identify the whistling animals is a non-invasive alternative to isolation conditions and cheaper than passive acoustic recording tags. Guiana dolphins can be observed during the entire year in Ilheus, Bahia State, Brazil, mainly in two areas at a distance of approximately 4 km from each other: An estuarine system named Pontal bay and Port of Ilheus (Izidoro and Le Pendu, 2012a). Seventy-eight dolphins have been photoidentified from 2007 to 2011 (Le Pendu et al., 2011). The first description of whistles of the Guiana dolphins from Ilheus was published in a comparative study of the whistle structure from ten different populations (Rossi-Santos and Podos, 2006). The authors recorded 24 whistles with frequency values ranging from 6 to 18 khz. The purpose of this work is to investigate the occurrence of signature whistles in the population of S. guianensis from Ilheus, northeastern Brazil, describing the acoustic structure of the whistles and by applying photo-identification methods to identify the individuals potentially emitting stereotypical whistles. II. MATERIALS AND METHODS A. Study area The municipality of Ilheus is situated on the southern coast of Bahia and is characterized by hot and humid tropical climate, with no dry season and monthly temperatures averaging between 21 Cand24 C(Andrade, 2003). The watershed includes four major rivers: Santana, Cachoeira, Fund~ao, and Almada and presents an extensive coverage of mangroves and small islands in the estuarine portion of the river. The rapid expansion of urban areas is a threat to estuaries and mangroves (Fidelman, 2001; Rodrigues and Martinez, 2014). B. Data collection Bioacoustics and photographic data were collected from a boat between October 2012 and July 2013, among latitudes S and S and longitudes W and W, i.e., between the coast and the 10 m bathymetric line. Most of the photo-identifications of dolphins were acquired by boat but complementary photo-identification sessions were conducted from a landed station at the edge of a breakwater, where Guiana dolphins can be closely observed (Izidoro and Le Pendu, 2012b; Santos et al., 2013). A 21 ft long fiberglass boat Travessia 21 (Krause boats, Arariu de Formiga, SC, Brazil), equipped with a 50 HP Mercury outboard engine (Mercury Marine, Fond du lac, WI), was used. Many data were collected near the extremity of the breakwater of Port of Ilheus (14 46 S and W), due to the regular occurrence of Guiana dolphins (Izidoro and Le Pendu, 2012a) (Fig. 1). All data were collected during good weather conditions (Beaufort 0 3). The Port of Ilheus is in a small cove with depth ranging between 2 and 10 m and receives cargo ships, cruise ships, and tug boats (Izidoro and Le Pendu, 2012a). Local fishing and recreational boats are also frequent in the area (Santos et al., 2013). Underwater sounds were recorded as wav files with three models of recorders: A Micro Track II (M-Audio, Cumberland, RI), a DR-07MKII (Tascam, TEAC Corporation, Tokyo, Japan), and a DR-100MKII (Tascam, TEAC Corporation, Tokyo, Japan), with 24 bits of precision at a sampling rate of 96 khz, coupled to a HTI-96MIN (High-Tech, Inc., Long Beach, MS) hydrophone with frequency response from 2 Hz to 30 khz. We sought to interfere as little as possible on animal behavior during recording. A group was initially approached at FIG. 1. Map of Ilheus coast, south of Bahia state, northeastern Brazil, showing the study area. J. Acoust. Soc. Am., Vol. 136, No. 6, December 2014 A. Lima and Y. Le Pendu: Signature whistles in Guiana dolphins 3179

3 a visually estimated distance of 50 to 100 m; however, sometimes dolphins came closer to the survey boat. We called recording occasion the set of whistles recorded from a group. Each recording occasion was carried out without any other dolphin group in the proximity. It was not possible to identify the whistler in the field and each recording potentially contained whistles from all dolphins in a given group. Photographs of dorsal fins were taken to identify group members, using one or two digital cameras: A Canon (Tokyo, Japan) EOS Rebel XT 350D equipped with a Canon 300 mm lens and a Canon EOS 7D, equipped with a 400 mm lens. There were at least three observers on board: A pilot, a recorder of sound emissions, and one or two photographers. One of the observers (AL) registered the following data on a standardized sheet: Date, beginning and end times of each group sighting, composition of the group (number of calves and adults), number of frames, and number of audio files. Individuals were identified as calves and adults, according to color visually estimated pattern and body size. Calves were smaller than adults and always accompanied by at least one adult. Calves had a gray-pinkish color while adult bodies had a more defined gray color (Randi et al., 2008; Santos et al., 2013). C. Data analysis 1. Spectograms Spectrograms (Hann window, overlap of 50%, with a discrete Fourier transform size of 256 points and 375 Hz) of the recorded whistles were created and analyzed using Raven Pro 1.5 (Cornell University). All recordings went through a primary visual spectrogram inspection in order to identify whistles and to determine quantitative sound parameters. We described the following frequency parameters (khz): Start, end, minimum and maximum frequency at 1 4, 1 2, and 3 4 of duration, as well as the frequency range. Duration and number of harmonics were also calculated. Whistle contours were categorized as ascending, descending, ascending descending, descending ascending, and others (see Monteiro-Filho and Monteiro, 2001; Azevedo and Van Sluys, 2005; Figueiredo and Sim~ao, 2009; May-Collado and Wartzok, 2009). Minimum values, maximum values, mean, standard deviation (SD), and frequency range values of whistle parameters were compared with other populations using published data. We used the XLSTAT s statistical analysis software 11 (addinsoft SARL) to conduct a Component Principal Analysis (PCA) and K-means cluster analysis in order to check if the mean parameters of frequency (start, end, minimum and maximum frequency at 1 4, 1 2, and 3 4 of duration) and duration were associated to population differences in whistles. The parameters of the Guiana dolphin population from Ilheus and another 11 localities (from South to North) were analyzed: North Bay/SC, Guaraqueçaba/PR, Cananeia/SP, and Puruba/SP (Azevedo and Van Sluys, 2005); Paraty Bay/RJ, SepetibaBay/RJ, and Guanabara Bay/RJ (Andrade et al., 2014); Fortaleza/CE; Marapanim/PA and Pacaja River/PA (Azevedo and Van Sluys, 2005) and Gandoca-Manzanillo in Costa Rica (May-Collado and Wartzok, 2009). Guanabara Bay was included as supplementary data in the PCA to enhance the clarity of the graph because it presented the highest values of frequency at 1 2 and 3 4 of duration, final, and maximum frequencies. We searched for whistles with stereotyped loops in sequences to identify potential signature whistles. A loop is a fundamental unit of a whistle that is repeated one or more times. Every time that condition was found, the spectrogram image was saved and identified with a code number (Fig. 2). We considered loop sequences those with very stereotyped contour (contours with great visual similarity) and an interval between successive whistles of less than 10 s (Figueiredo and Sim~ao, 2009; Janik et al., 2013). We analyzed each of the selected stereotypic whistles with respect to the number of loops and we measured on the first loop the following parameters: Duration, initial and final frequencies, minimum and maximum frequencies, number of inflection points (i.e., change in the direction of frequency modulation (ascending to descending and vice versa); and time interval between loops. 2. Photographs Photographs were selected by applying the methodology proposed by the Sarasota Dolphin Research Program (Sarasota Dolphin Research Program, 2006). We used both pictures from boat surveys and from complementary fixed point sessions in the photo-identification process. Only high quality photographs (good focus and perpendicular angle) with recognizable individuals, with a very distinct dorsal fin (at least two notches on the edges) were included. A smaller and unmarked animal seen next to a larger animal was assumed to be a calf. These and other not distinctive individuals were accounted, but their re-sighting could not be confirmed. Each good quality photograph of a dorsal fin FIG. 2. Spectrogram images of three whistles with stereotyped loops of Guiana dolphins in Ilheus and their code number J. Acoust. Soc. Am., Vol. 136, No. 6, December 2014 A. Lima and Y. Le Pendu: Signature whistles in Guiana dolphins

4 possessing marks was compared with an existing catalog of 80 previously photo-identified individuals (Le Pendu et al., 2011). A new photo-identified individual was eventually added to the catalog by selecting the best picture of its dorsal fin, considering focus, light, and sharpness. 3. Classification of whistles according to the visual inspection method We selected 68 stereotyped whistles that presented good visualization (with low interference from background noise) to apply the visual inspection method by external judges (Sayigh et al., 1995). Six judges with previous experience with sound analysis, but not with dolphin whistles, received printed spectrograms of the 68 selected whistles in a random presentation. They were asked to classify the whistles into assemblages according to the visual perception of the loop contours. An assemblage is a set of whistles formed by human judges through visual inspection. No information was given about the number of groups recorded, neither of emission contexts, as suggested by Nakahara and Miyazaki (2011). Each judge could create as many categories as they wanted. An association matrix between the whistles was built based on the number of times they were classified in the same assemblage by the judges. We hierarchically classified the whistle associations made by judges using a clustering method, using unweighted average clustering. This method forms clusters by pairs of pre-formed groups, based on the arithmetic average weight between the individual members and merge all the clusters that fall into a tie into a unique supercluster. We used Multidendrograms to visualize the clusters and the degree of heterogeneity inside a cluster (Fernandez and Gomez, 2008). Since original data were integers (0 to 6), the number of decimal significant digits for calculations was reduced to zero to obtain integer values of associations. 4. Combined analysis of sound recordings and photographs The whistles grouped by human judges were analyzed case by case, with respect to individuals present at each recording occasion. We sought for relations between the assemblages made by judges with photo-identified dolphins in each recording occasion. III. RESULTS A. Sampling effort From October 2012 to July 2013, we conducted 88 h of sampling effort in 30 days. We sighted 55 groups, totaling 210 dolphins (13.81% calves). The group size ranged from 1 to 8 (mean ¼ 3.818, SD ¼ 1.925). A total of 39 individuals (18 new) were identified with good-quality photographs. We sighted 55 groups in 30 days of field work and recorded the sound emissions from 21 groups on 9 different days, totaling 7 h of sound recording. B. Spectrograms The starting frequency, frequencies at first, 1 4, 1 2, and 3 4 of duration and final frequency of the 847 recorded whistles were similar to those reported in 10 other Brazilian and Costa Rican populations. Whistles were mainly ascending (59%; n ¼ 497), similar to what has been reported in other populations (from 57% in Marapanim to 97% in Guaraqueçaba). The fundamental frequency of the whistles recorded in Ilheus presented a minimum value of 0.8 khz and a maximum of 30.0 khz. The mean duration of a whistle ( ms) was shorter in Ilheus than in other Brazilian populations but similar to the Costa Rican population ( ms) (Table I). The first two principal components of the PCA accounted for 52.75% and 36.89% of the variance. Duration was poorly correlated with frequency variables. The first component F1 was highly correlated with frequency at 1 4 of duration (r ¼ 0.984) and minimum frequency (r ¼ 0.968), and both variables were highly correlated (r ¼ 0.975). The second component F2 was highly correlated with end r ¼ and maximum (r ¼ 0.936) frequencies, and both variables were also highly correlated (r ¼ 0.953). The graphic representation of PCA (Fig. 3) suggests that the 12 Guiana dolphin TABLE I. A comparative table of whistle mean frequency values (khz), percentage of ascending and mean duration (ms) from this study and ten other populations of Guiana dolphins arranged from south to north. StartF ¼ start frequency; 1 4 f ¼ frequency at 1 4 of duration; 1 2 f ¼ frequency at half of duration; 3 4 f ¼ frequency at 3 4 of duration; EndF ¼ final frequency. 1 ¼ North Bay/SC; 2 ¼ Guaraqueçaba/PR; 3 ¼ Cananeia/SP; 4 ¼ Puruba/SP (Azevedo and Van Sluys, 2005); 5 ¼ Paraty Bay/RJ; 6 ¼ SepetibaBay/RJ; 7 ¼ Guanabara Bay/RJ (Andrade et al., 2014); 8 ¼ This study; 9 ¼ Fortaleza/CE; 10 ¼ Marapanim/PA; 11 ¼ Pacaja River/PA (Azevedo and Van Sluys, 2005); 12 ¼ Gandoca-Manzanillo/Costa Rica (May-Collado and Wartzok, 2009). 1 Code Population MinF MaxF StartF 4 f 1 2 f 3 4 f EndF % ascending Duration (ms) 1 North Bay/SC % Guaraqueçaba/PR % Cananeia/SP % Puruba/SP % Paraty Bay/RJ % Sepetiba Bay/RJ % Guanabara Bay/RJ % Ilheus/BA % Fortaleza/CE % Marapanim/PA % Pacaja River/PA % Gandoca-Manzanillo/Costa Rica % J. Acoust. Soc. Am., Vol. 136, No. 6, December 2014 A. Lima and Y. Le Pendu: Signature whistles in Guiana dolphins 3181

5 FIG. 3. Principal component analysis using mean values of frequency and duration from 12 populations of Guiana dolphins. Factor 1 explains 52.75% of variance and factor 2 explains 36.89% of variance. Guanabara Bay/RJ (open circle) was included as supplementary data. 1 ¼ North Bay/SC; 2 ¼ Guaraqueçaba/PR; 3 ¼ Cananeia/SP; 4 ¼ Puruba/SP; 5 ¼ Paraty Bay/RJ; 6 ¼ SepetibaBay/RJ; 7 ¼ Guanabara Bay/RJ; 8 ¼ Ilheus; 9 ¼ Fortaleza/CE; 10 ¼ Marapanim/PA; 11 ¼ Pacaja River/; 12 ¼ Gandoca-Manzanillo/Costa Rica. populations form 3 distinct groups: The southern Brazilian populations on the left side of the first axis (1 to 4), the 4 northern populations with high minimum and initial frequencies on the right side of the first axis (9 to 12), and the 3 geographically intermediate populations (5 to 7) producing high ending and maximum frequency whistles situated above the second axis. The midmost position of Ilheus (8) on the graph is due to intermediate frequency values (see Table I). C. Combined analysis of visual inspection method and photographs Spectrograms of 68 stereotypic whistles recorded from 11 different groups were selected by spectrogram inspection for visual quantitative classification. Four pairs and 2 trios of whistles were assembled by all judges and 64 pairs of whistles were assembled by 5 judges (Fig. 4). Three out of the 6 assemblages formed by the 6 judges ({36, 44}, {14, 42}, and {31, 33, 41}) contained only whistles recorded on the same occasion and may have been produced by the same dolphin(s). The other three assemblages ({49, 55, 82}, {6, 73}, and {9, 68}) were whistles recorded on different occasions. Seven of the 13 assemblages formed by 5 judges were composed of whistles recorded on different occasions. A case-by-case analysis of the assemblages containing whistles recorded on different occasions provided clues about the identity of the whistlers (Table II). The identification of all the individuals in a group was not always possible with the motor switched off because the dolphins rarely remain close to the survey boat for a sufficient time and it was common to lose some individuals because of their dives and fast swimming velocity. In three cases, the assemblages formed by at least five judges included individuals identified on more than one occasion: Case 1: {75, 77} assemblage defined by 5 human judges. These whistles were recorded on 2 different field days, but dolphins D40 and D100 were present on both occasions (75 was recorded in a group of 7 individuals, but only D40 and D100 were identified). Whistle 77 was recorded in a group of 8 individuals and 5 were identified. Both whistles 75 and 77 may belong to D40 or D100, or to any other individual present in the groups in both occasions. Case 2: {49, 55, 82} assemblage defined by all 6 human judges. Whistles 49, 55, and 82 were recorded on different field days in groups of 3 (none identified), 6 (all identified), and 8 (5 identified), respectively. During the recording occasions of 55 and 82, D40, D100, and D101 were identified on both occasions. Case 3: {6, 73, 84}, assemblage defined by 5 human judges (with {6, 73} defined by all 6 human judges). The three whistles were recorded on different field days. When whistle 6 was recorded, the group of dolphins was not in sight but the recorded whistles were probably produced by 1 or 2 individuals because of the low number of whistles heard. Whistle 73 was recorded in a group of 5 individuals (all identified). Whistle 84 was recorded in a group of 8 individuals (5 identified). On the recording occasions of 84 and 73, the same dolphin individual (D105) was present. Whistles 77 (case 1), 82 (case 2), and 84 (case 3) were recorded on the same occasion in a group of 8 individuals, of FIG. 4. Multidendrogram of a hierarchical cluster analysis of the 68 stereotypic whistles, using an unweighted variable-group method using average clustering, with a co-phenetic correlation coefficient of The code number of the whistles assembled by all judges is in brackets. Whistles 77 and 84 are commented on in the text. The lower and upper bounds of gray rectangles indicate the minimum and maximum number of judges that associated whistles in a cluster J. Acoust. Soc. Am., Vol. 136, No. 6, December 2014 A. Lima and Y. Le Pendu: Signature whistles in Guiana dolphins

6 TABLE II. Assemblages made by at least five judges with identified dolphins present more than one recording occasion (in bold). Assemblage Number of judges Whistle code Group size Dolphins photo-identified I None 55 6 D40, D14, D24, D99, D100, D D80, D40, D101, D100, D105 II D03, D102, D103, D104, D D80, D40, D101, D100, D105 III D40, D D80, D40, D101, D100, D105 IV None 55 6 D40, D14, D24, D99, D100, D D80, D40, D101, D100, D D80, D40, D101, D100, D105 V 5 6 Unknown None 73 5 D03, D102, D103, D104, D D80, D40, D101, D100, D105 which 5 were identified: D40, D80, D100, D101, and D105. They were grouped in different assemblages by human judges (see Fig. 4) and were probably produced by different individuals. IV. DISCUSSION The whistles produced by Guiana dolphins in Ilheus are similar in frequency parameters and duration to those reported in other populations. Andrade et al. (2014) reported higher frequency range values on the fundamental component in three areas of southeastern Brazil [both minimum and maximum frequencies: 1.03 (minimum, in Paraty Bay) khz (maximum, in Sepetiba Bay)] compared to those reported in other areas; this probably occurred because the authors used recording equipment with higher upper frequency limits. Rossi-Santos and Podos (2006) found that starting, minimum frequencies, and duration of the whistles are more similar between adjacent sites than between distant populations. The principal component analysis discriminated northern and southern populations by frequency characteristics: Starting and minimum frequencies of the whistles were lower in southern populations. The starting frequencies of the six populations located south of Ilheus were lower than the same parameter for Ilheus. Similarly, frequencies at the 1 4, 1 2, and 3 4 of duration of all populations located north of Ilheus were higher than the same parameters for Ilheus. The upper position of Guanabara, Sepetiba, and Paraty Bay is due to higher ending and maximum frequencies and may be a consequence of the sampling frequency of 96 khz (vs a sampling rate of 48 khz used for the other populations, except the Costa Rican). Sepetiba and Paraty Bay populations were located with the other southeastern populations on the left side of the first axis. Our results confirm the findings of Rossi-Santos and Podos (2006) that latitudinal proximity indicates similarities in the frequency values of whistles. Conversely, Deconto and Monteiro-Filho (2013) described whistle parameters in Cananeia and Guaraqueçaba similar to those found in Costa Rica, and suggested that latitudinal variations in whistles reflect local adaptations to the environment. Andrade et al. (2014) found whistle variations between three localities that may also reflect an adaptation to ecological conditions: The higher frequency values recorded in the Guanabara Bay population may be an adaptation to louder ambient noise in that area. Following the definition of Janik and Sayigh (2013) that a signature whistle is a learned, individually distinctive whistle type in a dolphin s repertoire that broadcasts the identity of the whistle owner, the spectrograms analyzed suggest that Guiana dolphin whistles present particular frequency modulation patterns or contours, recognized by human judges. We identified modulation frequency patterns that served as a basis for a visual method of discrimination between whistles. The quantitative method has been criticized because it normalizes the duration of the whistles and then takes equally spaced frequency measurements of each whistle, sometimes failing to group similar contours (Sayigh et al., 2007). According to Janik (1999), the visual classification method is more reliable in identifying signature whistles. The use of a quantitative method, like the McCowan method (McCowan, 1995; McCowan and Reiss, 1995) or other, may be desirable to employ applicable calculations, and to categorize whistles with similar contours in spite of different durations, for example. The McCowan method assumes that the duration is irrelevant to the classification of whistles, but this is unlikely to be true. It is known that whistle duration is important in discriminating populations and behaviors in dolphin species and may also be an indicator of individual differentiation (May-Collado, 2013). Most studies about signature whistles involve captive animals (McCowan and Reiss, 1995) or capture-release events (Sayigh et al., 1999; Miksis et al., 2002; Janik et al., 2006), allowing an easy identification of the dolphin whistling. Here we used the photo-identification technique, trying to interfere as little as possible in the dolphins behavior. We could not photo-identify all the individuals, nor identify all the whistlers. Nonetheless, the presence of the same individuals in different recording occasions of stereotyped whistles (potential signatures) allowed us to infer an association between some types of whistles and the presence of specific individuals. The assemblages between similar whistles suggest that there are different types of whistles and they may J. Acoust. Soc. Am., Vol. 136, No. 6, December 2014 A. Lima and Y. Le Pendu: Signature whistles in Guiana dolphins 3183

7 have been emitted by the same individual. The smaller the group size, the greater the probability of assignment of signature whistles to individuals. Recording isolated individuals would confirm the existence of signature whistle in this population, but S. guianensis are rarely found alone in Ilheus (Izidoro and Le Pendu, 2012a). The relatively low number of whistles selected for classification (only 68 of 847 recorded) was expected because the animals produce more signature whistles when isolated from others (Janik and Slater, 1998). In addition, various whistles were excluded from analysis because of the interference of background noise. This study is the first that simultaneously uses both techniques of photo-identification and whistle recordings in S. guianensis. The results from the visual classification of stereotyped whistles corroborate the conclusions by Figueiredo and Sim~ao (2009), supporting the premise that S. guianensis may produce signature whistles. However, a prolonged sampling effort using the photo-identification technique and more than one hydrophone to increase localization potential of recorded groups (Quick et al., 2008) could increase the odds of identifying all the whistlers in a group. An adaptation of the SIGID method proposed by Janik et al. (2013) for T. truncatus to discriminate signature and non-signature whistles may also be applicable to S. guianensis. The application of these innovative methods should allow a definitive conclusion on the existence of signature whistles in S. guianensis. ACKNOWLEDGMENTS This study was granted by Cetacean Society International, Animal Behavior Society, and Universidade Estadual de Santa Cruz (Project No. PROPP ). The Fundaç~ao de Amparo a Pesquisa do Estado da Bahia (Fapesb Brazil) granted a master s degree scholarship to A.L. (BOL0175/2012). Several people helped us with logistic and data collection, including A. P. Andrade, J. Sousa, L. Vaz, Y. Cruz, J. Lacerda, E. Froes, G. Lima, V. Lima, and A. Conti. Gaston Gine helped us with the map confection. We particularly thank Universidade Estadual de Santa Cruz, Programa de Pos-Graduaç~ao em Zoologia (UESC), and Marina Ilheus, which provided logistical support. We thank INFRAERO who provided meteorological data. Andrade, L. G., Lima, I. M. S., Macedo, H., Carvalho, R. R., Lailson-Brito, J., Flach, L., and Azevedo, A. (2014). Variation in Guiana dolphin (Sotalia guianensis) whistles: Using a broadband recording system to analyze acoustic parameters in three areas of southeastern Brazil, Acta Ethol. Online first article, DOI: /s Andrade, M. P. (2003). Ilheus, Passado e Presente (Ilheus, Past and Present) (Editus, Ilheus, BA), 144 pp. Antunes, R., Schulz, T., Gero, S., Whitehead, H., Gordon, J., and Rendell, L. (2011). Individually distinctive acoustic features in sperm whale codas, Anim. Behav. 81, Azevedo, A., de, F., and Van Sluys, M. (2005). Whistles of tucuxi dolphins (Sotalia fluviatilis) in Brazil: Comparisons among populations, J. Acoust. Soc. Am. 117, Buckstaff, K. (2004). Effects of watercraft noise on the acoustic behavior of bottlenose dolphins, Tursiops truncatus, in Sarasota Bay, Florida, Mar. Mammal Sci. 20, Caldwell, M. C., and Caldwell, D. K. (1965). Individualized whistle contours in bottle nosed dolphins (Tursiops truncatus), Nature 207, Caldwell, M. C., and Caldwell, D. K. (1968). Vocalization of naive captive dolphins in small groups, Science 159, Caldwell, M. C., and Caldwell, D. K. (1970). Statistical evidence for individual signature whistles in pacific whitesided dolphins, Lagenorhynchus obliquidens, Cetology 3, 1 9. Caldwell, M. C., Caldwell, D. K., and Miller, J. F. (1973). Statistical evidence for individual signature whistles in the spotted dolphin, Stenella plagiodon, Cetology 16, Connor, R. C. (2000). Group living in whales and dolphins, in Cetacean Societies: Field Studies of Dolphins and Whales, edited by J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead (University of Chicago Press, Chicago, IL), pp Cook, M. L. H., Sayigh, L. S., Blum, J. E., and Wells, R. S. (2004). Signature-whistle production in undisturbed free-ranging bottlenose dolphins (Tursiops truncatus), Proc. Biol. Sci. 271, Deconto, L. S., and Monteiro-Filho, E. L. (2013). High initial and minimum frequencies of Sotalia guianensis whistles in the southeast and south of Brazil, J. Acoust. Soc. Am. 134, Erber, C., and Sim~ao, S. M. (2004). Analysis of whistles produced by the Tucuxi dolphin Sotalia fluviatilis from Sepetiba Bay, Brazil, An. Acad. Bras. Cienc. 76, Fernandez, A., and Gomez, S. (2008). Solving non-uniqueness in agglomerative hierarchical clustering using multidendrograms, J. Classif. 25, Fidelman, P. I. J. (2001). Manguezais do rio Santana, Ilheus, Bahia: caracterizaç~ao do sistema ( River Santana s mangroves, Ilheus, Bahia: system characterization ), Rev. Estud. Ambient. 3, Figueiredo, L. D. de, and Sim~ao, S. M. (2009). Possible occurrence of signature whistles in a population of Sotalia guianensis (Cetacea, Delphinidae) living in Sepetiba Bay, Brazil, J. Acoust. Soc. Am. 126, Gridley, T., Cockcroft, V. G., Hawkins, E. R., Blewitt, M. L., Morisaka, T., and Janik, V. M. (2014). Signature whistles in free-ranging populations of Indo-Pacific bottlenose dolphins, Tursiops aduncus, Mar. Mammal Sci. 30, Herzing, D. L. (1996). Vocalizations and associated underwater behavior of free-ranging Atlantic spotted dolphins, Stenella frontalis, and bottlenose dolphins, Tursiops truncates, Aquat. Mamm. 22, Izidoro, F. B., and Le Pendu, Y. (2012a). Estuarine dolphins (Sotalia guianensis) (Van Beneden, 1864) (Cetacea: Delphinidae) in Porto de Ilheus, Brazil: Group characterization and response to ships, North. West. J. Zool. 8, Izidoro, F. B., and Le Pendu, Y. (2012b). Estuarine dolphins (Sotalia guianensis, Cetacea, Delphinidae) play at Porto de Ilheus harbor, Bahia, Brazil, Panam. J. Aquat. Sci. 7, Janik, V. M. (1999). Pitfalls in the categorization of behaviour: A comparison of dolphin whistle classification methods, Anim. Behav. 57, Janik, V. M. (2000). Whistle matching in wild bottlenose dolphins (Tursiops truncatus), Science 289(5483), Janik, V. M., King, S. L., Sayigh, L. S., and Wells, R. S. (2013). Identifying signature whistles from recordings of groups of unrestrained bottlenose dolphins (Tursiops truncatus), Mar. Mammal Sci. 29, Janik, V. M., and Sayigh, L. S. (2013). Communication in bottlenose dolphins: 50 years of signature whistle research, J. Comp. Physiol. A. 199, Janik, V. M., Sayigh, L. S., and Wells, R. S. (2006). Signature whistle shape conveys identity information to bottlenose dolphins, Proc. Natl. Acad. Sci. U.S.A. 103, Janik, V. M., and Slater, P. J. B. (1998). Context-specific use suggests that bottlenose dolphin signature whistles are cohesion calls, Anim. Behav. 56, King, S. L., Sayigh, L. S., Wells, R. S., Fellner, W., and Janik, V. M. (2013). Vocal copying of individually distinctive signature whistles in bottlenose dolphins, Proc. Biol. Sci. 280(1757), Le Pendu, Y., Morais, B. C., Paula, Y. C. De, and Assis, C. V. de (2011). Residence and site fidelity of Guiana dolphin, Sotalia guianensis (Van Beneden, 1864) in the port of Ilheus, Bahia, Brazil, Abstract. B: 19th Biennial Conference on the Biology of Marine Mammals, Tampa, FL, Soc. Mar. Mammal., pp May-Collado, L. J. (2013). Guyana dolphins (Sotalia guianensis) from Costa Rica emit whistles that vary with surface behaviors, J. Acoust. Soc. Am. 134, EL359 EL J. Acoust. Soc. Am., Vol. 136, No. 6, December 2014 A. Lima and Y. Le Pendu: Signature whistles in Guiana dolphins

8 May-Collado, L. J., Agnarsson, I., and Wartzok, D. (2007). Phylogenetic review of tonal sound production in whales in relation to sociality, BMC Evol. Biol. 7, 136. May-Collado, L. J., and Wartzok, D. (2008). A comparison of bottlenose dolphin whistles in the Atlantic ocean: Factors promoting whistle variation, J. Mammal. 89, May-Collado, L. J., and Wartzok, D. (2009). A characterization of Guyana dolphin (Sotalia guianensis) whistles from Costa Rica: the importance of broadband recording systems, J. Acoust. Soc. Am. 125, McCowan, B. (1995). A new quantitative technique for categorizing whistles using simulated signals and whistles from captive bottlenose dolphins (Delphinidae, Tursiops truncatus), Ethology 100, McCowan, B., and Reiss, D. (1995). Quantitative comparison of whistle repertoires from captive adult bottlenose dolphins (Delphinidae, Tursiops truncatus): A re-evaluation of the signature whistle hypothesis, J. Comp. Psychol. 109, Miksis, J. L., Tyack, P. L., and Bruck, J. R. (2002). Captive dolphins, Tursiops truncatus, develop signature whistles that match acoustic features of human-made model sounds, J. Acoust. Soc. Am. 112, Monteiro-Filho, E. L. D. A., and Monteiro, K. D. K. A. (2001). Low-frequency sounds emitted by Sotalia fluviatilis guianensis (Cetacea: Delphinidae) in an estuarine region in southeastern Brazil, Can. J. Zool. 79, Nakahara, F., and Miyazaki, N. (2011). Vocal exchanges of signature whistles in bottlenose dolphins (Tursiops truncatus), J. Ethol. 29, Parijs, S. M. Van, and Corkeron, P. J. (2001). Vocalizations and behaviour of Pacific Humpback Dolphins Sousa chinensis, Ethology 107, Pivari, D., and Rosso, S. (2005). Whistles of small groups of Sotalia fluviatilis during foraging behavior in southeastern Brazil, J. Acoust. Soc. Am. 118, Quick, N. J., Rendell, L. E., and Janik, V. M. (2008). A mobile acoustic localization system for the study of free-ranging dolphins during focal follows, Mar. Mammal Sci. 24, Randi, M. A. F., Rassolin, P., Rosas, F. C. W., and Monteiro-Filho, E. L. de A. (2008). Padr~ao de cor de pele ( Pattern of skin color ), in Biologia, Ecologia e Conservaç~ao do Boto-cinza (Biology, Ecology and Conservation of the Guiana Dolphin), edited by E. L. A. Monteiro-Filho and K. D. K. A. Monteiro (C^amara Brasileira do Livro, S~ao Paulo), pp Rodrigues, N. N., and Martinez, R. A. (2014). Wildlife in our backyard: Interactions between Wied s marmoset Callithrix kuhlii (Primates: Callithrichidae) and residents of Ilheus, Bahia, Brazil, Wildlife Biol. 20, Rossi-Santos, M. R., and Podos, J. (2006). Latitudinal variation in whistle structure of the estuarine dolphin Sotalia guianensis, Behaviour 143, Santos, M. S., Schiavetti, A., and Alvarez, M. R. (2013). Surface patterns of Sotalia guianensis (Cetacea: Delphinidae) in the presence of boats in Port of Malhado, Ilheus, Bahia, Brazil, Lat. Am. J. Aquat. Res. 41, Sarasota Dolphin Research Program (2006). Manual for Field Research and Laboratory Activities (Sarasota, FL), 60 pp. Sayigh, L. S., Esch, H. C., Wells, R. S., and Janik, V. M. (2007). Facts about signature whistles of bottlenose dolphins, Tursiops truncates, Anim. Behav. 74, Sayigh, L. S., Tyack, P. L., Wells, R. S., Scott, M. D., and Irvine, A. B. (1995). Sex difference in signature whistle production of free-ranging bottlenose dolphins, Tursiops truncates, Behav. Ecol. Sociobiol. 36, Sayigh, L. S., Tyack, P. L., Wells, R. S., Solow, A. A. R., Scott, M. D., and Irvine, A. B. (1999). Individual recognition in wild bottlenose dolphins: A field test using playback experiments, Anim. Behav. 57, Shapiro, A. D. (2006). Preliminary evidence for signature vocalizations among free-ranging narwhals (Monodon monoceros), J. Acoust. Soc. Am. 120, J. Acoust. Soc. Am., Vol. 136, No. 6, December 2014 A. Lima and Y. Le Pendu: Signature whistles in Guiana dolphins 3185