Bird vocalizations might be used for specific recognition, territorial defense, and reproduction. Bioacoustic studies aim to understand the production, propagation and reception of acoustic signals, and they are an important component of research on animal behavior and evolution. In this study we analyzed the sound structure of duet vocalizations in pairs of Myiothlypis leucoblephara and evaluated whether the vocal variables differ among pairs and if there are differences in temporal characteristics and frequency of duets between pairs in forest edges vs. forest interior. Vocalizations were recorded from 17 bird pairs in three remnants of Atlantic Forest in southern Brazil. Six of the bird pairs were situated at the edge of the forest remnant, and 11 were in the interior of the remnant. The duets of different pairs between forest areas showed descriptive differences in the frequency, number of notes per call, and time between issuance of calls, with the main distinguishing feature being a change in frequency of a few notes in the second part of the musical phrase. The minimum frequency of vocalization was reduced at the private area than in the other two remnants (p <0.05). The duets of birds in the forest edge and forest interior did not significantly differ in minimum or maximum frequency of phrases (p> 0.05), phrase duration (p> 0.05) or number of notes per phrase (p> 0.05). Myiothlypis leucoblephara did not show a specific pattern with respect to issue of phrases in duets, but instead showed five different patterns, which were variable among pairs. There was a sharp decline or alternation in frequency between notes in the second part of the musical phrase for recognition among pairs. Variation in vocalization among M. leucoblephara duets may play a role in pair recognition.

A vocalização em aves pode ser usada para reconhecimento de espécies, defesa territorial e reprodução. Estudos bioacústicos têm o objetivo de entender a produção, propagação e recepção de sinais acústicos, sendo importantes em pesquisas de comportamento animal e evolução. No presente estudo, analisamos a estrutura sonora da vocalização em duetos de Myiothlypis leucoblephara e avaliamos se as variáveis ​​vocais diferem entre os pares e se existem diferenças nas características temporais e na frequência de duetos entre pares nas bordas da floresta versus interior da floresta. Vocalizações foram registradas de 17 pares de aves em três remanescentes florestais no oeste de Santa Catarina, no sul do Brasil. Seis dos pares de aves estavam situados na borda do remanescente da floresta e 11 no interior. Utilizamos os testes de Kruskal-Wallis e Mann-Whitney U. Os duetos de pares diferentes analisados em todos os remanescentes apresentaram diferenças descritivas na frequência, número de notas por frase e tempo na emissão de notas, sendo que a principal característica é a mudança de frequência de algumas notas da segunda parte da frase. A frequência mínima das vocalizações foi menor na área da Aurora (p<0,05). Os duetos comparados entre borda e interior de mata não tiveram diferenças significativas de frequência mínima das frases (p > 0,05), frequência máxima (p > 0,05), tempo de duração das frases (p > 0,05) e número de notas por frase (p > 0,05). A espécie não apresentou um padrão específico na emissão das frases nos duetos, apresentando cinco padrões diferentes e estes são variáveis entre os pares. Myiothlypis leucoblephara apresenta características de diminuição brusca ou alternância de frequência entre notas da segunda parte da frase. Estas variações na vocalização em diferentes duetos da espécie apontam as características de reconhecimento entre os pares.

Atlantic forest, bioacoustics, vocal behavior

Acomportamento vocal, bioacústica, Floresta Atlântica

Vocal communication is the exchange of information made between sender and receiver individuals (Vielliard 2000) and is an important way of information exchange between birds (Sick 1997; Silva and Vielliard 2011). Bioacoustic studies examine animal vocal displays (Sigrist 2006) and aim to understand the production, propagation and reception of acoustic signals. Such studies are an important component of research into animal behavior and evolution (Salvador 2008).

Bird vocalizations are mainly important for territory defense, attraction of the opposite sex (Vielliard 1987; Sick 1997; Storer et al. 2002; Sigrist 2006; Pijanowski et al. 2011; Pereira 2011) and species recognition (Vielliard 1987; Silva 1995), and play a key role in the evolution and survival of species and ecosystem function (Vielliard 1987). Bird acoustic signals are essential for communication (Vielliard 2000), and vocalization height and amplitude variables can be related to home range location and spatial distance between individuals (Storer et al. 2002).

The acoustical code allows for the recognition of individuals of the species. Individual recognition can be achieved if an individual emits an acoustic signal distinct from the signal of other individuals, even though the specific recognition code might still be retained (Vielliard 2004). It is well understood that birds might recognize each other through vocalization, but little is known about the specific features of the vocalizations that facilitate individual recognition in territorial birds (Aubin et al. 2004).

Some birds may vocalize in duets, which occurs when two conspecific individuals vocalize in combination with each other either synchronously or alternately, usually in response to reproductive interaction or territory defense (Langmore 1998). The duet has the function of cooperative protection of territory, as well as territorial care in response to reproduction and paternal custody during the fertility of the females (Topp and Mennill 2008).

Myiothlypis leucoblephara (Vieillot, 1817), commonly known as the white-browed-warbler, is a common and abundant bird in the Atlantic forest (Aubin et al. 2004). This species is typical of underbrush and montane forests (Sigrist 2006) and maintains intense vocal activity throughout the year (Silva 1991). Its vocalization is strong and melodious, starts quickly and goes on a clear and descending scale (Develey 2004), consisting of a succession of very similar notes that slowly decrease in frequency (Aubin et al. 2004). Aubin et al. (2004) produced the only study on M. leucoblephara vocalization, in which individual vocalizations were evaluated. No studies, to our knowledge, have examined duet vocalizations in this species.

In this study we analyze duet vocalizations’ structure in M. leucoblephara, seeking to determine song characteristics involved in individual recognition between pairs. We also compare the physical variables of song units between duetting birds occupying two different environments: forest edges and interior. We work with the hypothesis that there are duets with different characteristics, which may indicate the recognition between the individuals of a pair.

We expected to record higher and lower frequencies in Hz and a larger number of notes per phrase in the edge environment, because this site has more noise interference when compared to the interior of the forest.

The phrases of all duets are structured by emission of notes of descending modulation. There are slight frequency changes between notes in different duets in the middle of the phrase (completion notes), when there is alternating low and high frequency of some notes and a sudden drop in frequency in others (Fig. 2), and the characteristic is the same for pairs within a duet, but not the same for different pairs. The acoustic characteristics of the sentences are the same for the two individuals in the duet, but different for different pair of duets.

Figure 2. 

Sonogram of a duet of M. leucoblephara registered in PA (A) and NF-1 (B), in southern Brazil. Two individuals appear vocalizing in a duet, one after the other. Sonogram of the complete vocalization sentence of M. leucoblephara (C). NF-1 = fragment in the National Forest of Chapecó; PA = private area.

Vocalizations were more frequent between 7 am and 10 am, and near dusk. As the popular name of the bird suggests, vocalizations occur while the bird hops forward along twigs and branches of the forest understory. Myiothlypis leucoblephara vocalization consists of one phrase that initiates with high frequency and noticeably decreases throughout the duration of the song (Fig. 2. The general classification of the notes emitted in the phrase for all individuals was acute tone with pure acoustic signals (only with a frequency) and downward modulation. The call consists of a single note. We found differences in the number of notes, timing and frequency of phrases of duets between different bird pairs.

The total number of notes in the phrases ranged from 13 to 32, and full phrase duration ranged from 3.074 to 8.325 seconds (mean = 5.243, SD = 1.396). The minimum frequency ranged between 2.566 to 5.037 Hz (mean = 3.397, SD = 609.7) while maximum frequency between 7.318 and 10.075 Hz (mean = 8.803 SD = 854.7) (Suppl. material 1). The minimum frequency of the first note in the phrase ranged from 5.037 to 7.603 Hz (mean = 6.399, SD = 668.9) while the maximum ranged from 7.508 to 9.694 Hz (mean = 8.747, SD = 673.5). The minimum time between the first and second note of the phrase was 0.209 seconds, and the maximum time was 1.550 seconds (Suppl. material 2).

Due to the distance of one of the individuals in a pair, in seven recordings (41.2%) some of the initial, high frequency notes were faded or even imperceptible on the spectrogram. This (notes faded or imperceptible on the spectrogram) was not observed from the middle of the phrase and beyond, as these notes tend to have lower frequency and thus have better clarity on the spectrogram.

Birds from private area site have lower minimum frequencies than NF 01 and NF 02 (KW = 13.572, df = 2, P < 0.05). NF 01 and 02 did not differ in minimum frequency (P > 0.05), and there were no differences among the three forest remnants in maximum frequency (KW = 5.433, df = 2, P > 0.05) or phrase duration (KW = 0.551, df = 2, P > 0.05) (Fig. 3).

Figure 3. 

Mean and SD of minimum and maximum frequency (Hz) and phrase duration (s) in M. leucoblephara duets between November 2014 and May 2015 in three forest remnants (PA, NF-1 and NF-2) located in southern Brazil. NF-1 and NF-2 = fragments in the National Forest of Chapecó; PA = private area.

There were no differences in phrase minimum (U = 81.000, df = 1, P > 0.05), maximum frequency (U = 72.500, df = 1, P > 0.05), phrase duration (U = 77.000, df = 1, P > 0.05) or number of notes per phrase (U = 60.000, df = 1, P > 0.05) between duets on the forest edge vs. the interior. The forest edge and interior did not differ in minimum frequency of the first note of each phrase (U = 44.500, df = 1, P > 0.05) nor the time (s) between the first and second note of each phrase (U = 44.500, df = 1, P > 0.05). The maximum frequency of the first note of each phrase was marginally higher in the forest interior (U = 21.000, df = 1, P = 0.049).

We discovered five different patterns of overlapping phrases in duets, with one pair demonstrating two vocalization patterns and the other pairs showing only one pattern (Fig. 4). We call the first patterns ‘p1’, which was present in 12 different duets (70.5%). In this pattern the first individual starts and ends the phrase, then the other individual vocalizes. We call the second pattern ‘p2’, which was recorded in two different duets (11.7%). Other vocalization patterns were recorded only once. Pattern 1 (p1): one individual vocalizes after the other, with no overlapping phrases or notes between the pair. In this pattern no time elapses between emission of phrases by the first individual and the reply from the other individual.

Figure 4. 

General vocalization patterns. NI = introductory notes (rectangles with finer lines); NF = final notes (rectangle thicker lines); 1 = individual 1; 2 = individual 2; p1, p2, p3, p4 and p5 = patterns 1, 2, 3, 4, and 5, respectively.

Pattern 2 (p2): the first individual emits introductory notes, and prior to final notes, individual 2 begins a phrase and they vocalize synchronously for a period of time. In this period individual 1 emits final notes while individual 2 emits introductory notes, then emits final notes alone.

Pattern 3 (p3): the pair vocalizes synchronously.

Pattern 4 (p4): individual 1 emits the whole phrase, but in the final part of the phrase (final notes) individual 2 begins a phrase consisting only of introductory notes and stopping shortly thereafter. The pair ends the phrase together, sometimes emitting different portions. Pattern 5 (p5): individual 1 begins introductory notes and the other individual emits the final notes. Thus, each individual contributes by vocalizing one part of the duet.

The only pattern recorded in all three remnants (PA, NF 01 and 02) and in both forest areas (edge and interior) was pattern 1. Pattern 2 was recorded only in NF 01, both in the forest interior and edge. Pattern 3 was recorded only in the interior of NF 01. Pattern 4 was only recorded in PA, and pattern 5 was recorded in the forest interior of NF 01. NF 02 was the only remnant with no recordings with phrase overlapping in duets.

The differences between M. leucoblephara pairs among forest remnants in duet characteristics, such as note frequency range and number of notes per phrase, are likely indicators of family group recognition. These birds have the sensitivity to recognize vocal alterations in notes, increase in the number of notes, and frequency variations, which they use to identify partners, family members, and invaders in their territory (Carvalho 2010). The various functions of avian duet vocalization are diverse and subject to controversy, but may also involve individual recognition (Vielliard 1987).

Individual recognition is generally associated with species that have similar vocalizations as in M. leucoblephara, suggesting that species with individual vocal repertoires are sensitive to subtle differences in vocalization (Aubin et al. 2004). Vocalization structure can be diverse. Acoustic elements can be shared but issued differently; different acoustic elements may be emitted in sequence, or imitation may occur without a defined sequence. These categories are not found in isolation, as individuals from the same species may fit into different categories (Vielliard 2004).

Volatinia jacarina (blue-black-grassquit) (Thraupidade, Passeriformes) and M. leucoblephara showed a huge inter-individual variation in the physical structure of vocalization. Individuals can accurately repeat extremely complex details of a song which lasts around half a second, but identical structures are not found among different individuals (Vielliard 2004). Five Turdus species (Turdidae, Passeriformes) show significant overall differences between individuals in both the temporal and frequency variables of vocalizations (Nascimento 2012). Troglodytidae and Turdidae have a minimum number of within-species differences in vocalization patterns and individual recognition information is thus reduced to characteristics of note rhythm or harmonic structure, demonstrating intra-individual versatility (Vielliard 1987). Oryzoborus angolensis (chestnut-bellied seed-finch) also showed significant inter-individual differences in note duration, time interval between notes, notes per second, and minimum and maximum frequency (Lopes 2011). The results from this study reinforce the concept that each species or individual uses different vocal recognition characteristics. The vocal characteristics of M. leucoblephara are shared between duet partners (i.e., with respect to the emission of phrases) and pairs can accurately mimic vocalization structure. One study of isolated individuals in captivity discovered potential clues about the identity of the individual signature in the first half of the phrase (Aubin et al. 2004). In our study we found small differences at the beginning of the phrase, but the main differences between units were found in the second half of the sentence.

Individual acoustic recognition was first discovered in the Laridae family, a seabird group that nests in dense colonies and cannot use visual location signals. They instead use harmonic variation in vocalization to identify partners, or to distinguish individuals within organized societies (i.e., as in Cacicus spp.) (Vielliard 1987). Penguins also perform duets for individual recognition when searching for partners among the multitude of individuals gathered during the incubation period (Vielliard 1987). Individual recognition is a prerequisite for social interactions, including parental care and cooperation between breeding pairs (Stoddard 1996).

Habitat loss in a steppe matrix markedly affected song-type sharing mechanisms in Chersophilus duponti (Dupont’s lark). In fragmented habitats, song sharing among neighbors was better, possibly because of harsher competition for limited resources, whereas sharing among non-neighbors was lower, probably because of the lack of interactions among individuals isolated by habitat barriers. Local habitat fragmentation may alter behavioral processes by hampering the local transmission of song cultures, in particular the spatial range of individual acoustic niche contracts at increasing fragmentation levels (Laiolo and Tella 2005). The PA, which differed from the others, is a smaller area and is not a protected area; it still suffers anthropic disturbance from the surrounding community.

Small variations in phrase timing and frequency among some duets are considered irrelevant, as birds are rarely able to vocalize with the exact same frequency and time variables as the other individual (Lima 1997). Moreover, the degree in precision of mimicking vocal characteristics can be influenced by environmental factors such as rain and wind, and at times equipment errors may limit the ability to accurately assess (Lima 1997).

Myiothlypis leucoblephara duet performance may be associated with territorial defense or food resource. Several tropical birds vocalize in duets throughout the year to defend territories (Langmore 1998). This territoriality is considered a selection strategy, wherein an animal yields possession of a location, food resource, or partner because another individual arrived first (Rose 2000).

Although the acoustic structure of M. leucoblephara is simple, there are propagation problems associated with the restrictive environment of a tropical forest and there is a need to transmit efficient information (common in species with specific recognition), the vocalization of the species satisfies an important requirement of territoriality: the identification between neighbors (Aubin et al. 2004). Analysis of duets revealed that each pair has unique characteristics regarding the issue of notes within phrases, which makes other individuals or intruders easily recognizable.

Most duet recordings had some missing or imperceptible notes in the spectrogram, always those with highest frequency in every phrase. Higher frequencies have more rapid attenuation than do low frequency in environments with obstacles such as branches, leaves, or air turbulence or viscosity (Krebs and Davies 1996). Experiments by Aubin et al. (2004) with the same species showed that beyond a distance of 100 m, the first notes are absent in some individuals, and hence, the vocal signature disappears in the background noise. According to the authors, short-range propagation of vocal information seems to be an adaption of territorial species, which serves to reduce aggressive responses to unknown birds and, hence, limit unwarranted energy consumption for territorial defense.

Morton (1975) showed that birds near the ground would be able to transmit acoustic signals further in the frequency range of 1000–2500 Hz, and that attenuation is increased above 2,000 Hz. Our study does not support this assessment, as M. leucoblephara lives in the understory and had minimum and maximum vocalization frequency between 2,566 and 10,075 Hz, respectively. Maybe the species has no need to communicate across long distances.

Myiothlypis leucoblephara showed no significant differences in vocalization structure between the forest edge and interior. A study by Silva (1991) concluded that this species does not show significant differences, suggesting that vocalization is adapted to a particular environment. These results contradict those found by other authors for most species. The structure of bird vocalization is closely related to the habitat in which they live, and this can be seen at the scale of forest edge, interior, or opening (Morton 1975; Storer et al. 2002). Habitats differ with respect to density and vegetation type, resulting in different selection pressures on acoustic signals, as a consequence of the way the sound is attenuated and degraded as it penetrates the physical environment (Slabbekoorn and Smith 2002). Environmental factors may select for specific vocalization characteristics, sometimes causing acoustic convergence among populations living in the same habitat and divergence among populations living in different habitats (Aubin et al. 2004). Other species in the Parulidae family, such as Myiothlypis flaveola (flavescent-warbler) and Myiothlypis leucophrys (white-striped-warbler), showed significant differences in maximum and minimum frequency, number of notes, and duration of vocalization between populations in urban vs. rural fragments (Carvalho 2010). Myiothlypis fulvicauda (buff-rumped-warbler) vocalizations differed based on environment and geographical variation (Escalante 2013).

Environmental and social factors are not enough to explain the diversity of bird vocalizations, because closely related species groups often have vocalizations with similar acoustic properties. Regardless, some vocalization characteristics evolve in response to the selection, while others remain similar. This may occur because vocalization characteristics are neutral with respect to the physical environment, or because evolutionary change is limited by genetic correlations or lack of variance (Van Buskirk 1997).

The birds can recognize neighbors or intruders in their territory through the detection of simple variations in vocal repertoire. Myiothlypis leucoblephara showed no significant differences in structural characteristics, minimum or maximum frequency, number of notes per phrase, or time of phrase execution between forest edge and interior environments. The vocalization of a species has a characteristic of its own (Silva 1991), not being modified by the type of habitat that lives.

Myiothlypis leucoblephara did not show species-specific duet patterns, instead presenting several emission patterns and overlapping phrases in the duets. Duet vocalization occurs when two individuals vocalize in combination with one another or in an alternating synchronized manner (Langmore 1998). The participation of each individual in the duet and organization of it are variable for each species or different individuals (Vielliard 1987). The duet may consist of verses that last for the same period of time but are not linked together; also one individual can vocalize after the other and the verses may be the same or different (Sick 1997).

Further studies should be carried out on duets of other species to evaluate syntax modification, phrase modulation, and dispersion of pairs or family members to assess the territory covered by groups. Playback experiments may also be useful to test the extent of individual recognition, as well as phylogenetic studies of vocalization in M. leucoblephara and other species of Myiothlypis.

We thank Elaine M. Lucas Gonsales and Ronei Baldissera for useful suggestions for the experiment and the manuscript. We thank the anonymous reviewers for their contributions to the manuscript. We thank the Laboratory of Ecology and Chemistry at the Universidade Comunitária da Região de Chapecó (Unochapecó) for loaning field materials. The Chico Mendes Institute for Biodiversity Conservation (ICMBio) provided the license for data collection at the National Forest of Chapecó.