Song Learning in Wild and Domesticated Strains of White‐Rumped Munia,Lonchura striata,Compared by Cross‐Fostering Procedures: Domestication Increases Song Variability by Decreasing Strain‐Specific Bias

Song diversity results from the interactions between natural selection, sexual selection, and individual learning. To understand song diversity, all three factors must be considered collectively, not separately. Bengalese Finches were domesticated about 250 yr ago. Their courtship songs have become different from their ancestor, the White‐rumped Munia. Bengalese Finches sing songs with complex note‐to‐note transition patterns and with acoustically diverse song notes while White‐rumped Munias sing songs with fixed note sequence and mostly broad band song notes. Bengalese Finches were selected for domestication based on their good parenting ability, not their songs, but this artificial selection has nonetheless affected their songs. To test whether divergence occurred not only in the song phenotypes but also in the genetic basis for predisposition of strain specific song learning, we conducted a cross‐fostering experiment between Bengalese Finches and White‐rumped Munias. In both strains, song learning was affected by rearing condition: the acoustical feature and transition patterns followed those of the foster fathers. However, the accuracy of song learning differed between the wild and the domesticated strains: sharing of song note between sons and tutors in Finches was not very accurate regardless of the tutor, while Munias were highly accurate in copying Munia songs but often omitted song elements from Finch fathers. These results suggest that White‐rumped Munias are strongly constrained to learn their own strain’s song, and that this constraint was relaxed in the Bengalese Finch by domestication.     

Song Learning in Bengalese Finches: a Comparison with Zebra Finches

Male Bengalese finches, Lonchura striata var. domestica, learn their song from an adult male conspecific with whom they can interact at 35 to 70 days of age and normally-raised males fail to reproduce song which they have only heard before or after this time. Birds which have been raised by their mother alone and those which have been deprived of a song tutor during the learning phase produce abnormal songs with indistinct elements and little or no phrase structure; this is typical of males which fail to hear adult song during their development. These songs are unstable and are replaced by normal songs, if there is an opportunity to learn from an adult male conspecific. Presumably, this flexibility in the time when young males learn acts as a safeguard to ensure that normal conspecific song is produced. These results bear striking similarity to those on zebra finch song development. Differences between the two species, especially in the learning of call notes by female zebra finches, are discussed.     

Diet and birdsong: short‐term nutritional enrichment improves songs of adult Bengalese finch males

Song is a notable sexual signal of birds, and serves as an honest indicator of male quality. Condition dependence of birdsong has been well examined from the viewpoint of the developmental stress hypothesis, which posits that complex songs assure fitness because learned acoustic features of songs are especially susceptible to early‐life stress that young birds experience in song learning periods. The effect of early stress on song phenotypes should be crucial, especially in age‐limited song learners which sing stereotyped songs throughout life. However, little attention has been paid to non‐learned song features that can change plastically even in adulthood of age‐limited song‐learners. Although it has been shown that food availability affects song rate in wild songbirds, there is limited evidence of the link between favorable nutritional conditions and song phenotypes other than song rate. Under the prediction that singing behavior reflects an individual's recent life history, we kept adult Bengalese finch males under high‐nutrition or normal diet for a short term, and examined changes in body mass and songs. We found that birds on a high‐nutrition diet showed higher song output (e.g. song rate and length) compared with those of the control group, while changes in body mass were moderate. In addition, note repertoire became more consistent and temporal structures got faster in both nutrition and control groups, which indicates that songs were subject to other factors than nutrition. Considering that female estrildid finches, including Bengalese and zebra finches, show a preference toward complex songs as well as longer songs and higher song rate, it is plausible that different aspects of singing behavior signal different male qualities, and provide multifaceted clues to females that choose mates.     

Comparative study of long-term reduction of high-frequency song syllables in deafened zebra and Bengalese finches: implications for the role of auditory feedback in song maintenance

Auditory feedback is necessary for adult song maintenance in both oscines and psittacines. Although belonging to phylogenically separated orders, deafened adult oscine Bengalese finches and psittacine budgerigars exhibit similarities in certain aspects of song changes. An interesting question is whether these birds share common mechanisms for song maintenance. Therefore, it is important to compare the effects of deafening on adult song patterns among and within orders. Although zebra and Bengalese finches are closely related oscine species, few studies have performed direct, long-term, quantitative comparisons of their songs after deafening because suitable song characteristics have not been identified. Based on our previous findings for Bengalese finch songs, we analyzed zebra finch songs over 9 months after deafening, focusing on changes in the number of syllables categorized according to fundamental frequencies. Deafened zebra finches demonstrated a gradual but significant decrease in high-frequency syllables and a tendency to increase low-frequency syllables, similar to deafened Bengalese finches. Although this change took longer in zebra finches, the altered proportion of syllables eventually stabilized. Results indicated that adult songs show similar aspects after auditory deprivation, and that neural mechanisms involved in the maintenance of high-frequency song syllables, using auditory feedback, may be present in both finches despite species differences.     

Song learning with audiovisual compound stimuli in zebra finches

We investigated the effects of audiovisual compound training on song learning in zebra finches, Taeniopygia guttata. In the first experiment, presentation of a stuffed adult zebra finch male was found to be reinforcing to zebra finch males in an operant task. In a separate experiment, zebra finch males were reared without their father from day 7 after hatching onwards. Between 35 and 76 days, they were placed in isolation and exposed to taped songs of a zebra finch male, according to a random schedule (20 presentations/h). For half of the birds, presentation of the song coincided with presentation of a stuffed zebra finch male. For the remaining birds, each presentation of the song was followed by presentation of a stuffed male. The birds were subsequently isolated until day 142, when their own songs were recorded and analysed. Birds in both groups shared significantly more song elements with their tutor songs than with an unfamiliar song. There was no significant difference in song learning between the groups. These results confirm that zebra finches can learn part of their songs from taped tutor songs. Furthermore, simultaneous presentation of the tutor song and a relevant, salient visual stimulus is not superior to sequential presentation. Copyright 1999 The Association for the Study of Animal Behaviour.     

Developmental experience alters information coding in auditory midbrain and forebrain neurons

In songbirds, species identity and developmental experience shape vocal behavior and behavioral responses to vocalizations. The interaction of species identity and developmental experience may also shape the coding properties of sensory neurons. We tested whether responses of auditory midbrain and forebrain neurons to songs differed between species and between groups of conspecific birds with different developmental exposure to song. We also compared responses of individual neurons to conspecific and heterospecific songs. Zebra and Bengalese finches that were raised and tutored by conspecific birds, and zebra finches that were cross‐tutored by Bengalese finches were studied. Single‐unit responses to zebra and Bengalese finch songs were recorded and analyzed by calculating mutual information (MI), response reliability, mean spike rate, fluctuations in time‐varying spike rate, distributions of time‐varying spike rates, and neural discrimination of individual songs. MI quantifies a response's capacity to encode information about a stimulus. In midbrain and forebrain neurons, MI was significantly higher in normal zebra finch neurons than in Bengalese finch and cross‐tutored zebra finch neurons, but not between Bengalese finch and cross‐tutored zebra finch neurons. Information rate differences were largely due to spike rate differences. MI did not differ between responses to conspecific and heterospecific songs. Therefore, neurons from normal zebra finches encoded more information about songs than did neurons from other birds, but conspecific and heterospecific songs were encoded equally. Neural discrimination of songs and MI were highly correlated. Results demonstrate that developmental exposure to vocalizations shapes the information coding properties of songbird auditory neurons. © 2009 Wiley Periodicals, Inc. Develop Neurobiol 70: 235–252, 2010.     

A neural network model for generating complex birdsong syntax

The singing behavior of songbirds has been investigated as a model of sequence learning and production. The song of the Bengalese finch, Lonchura striata var. domestica, is well described by a finite state automaton including a stochastic transition of the note sequence, which can be regarded as a higher-order Markov process. Focusing on the neural structure of songbirds, we propose a neural network model that generates higher-order Markov processes. The neurons in the robust nucleus of the archistriatum (RA) encode each note; they are activated by RA-projecting neurons in the HVC (used as a proper name). We hypothesize that the same note included in different chunks is encoded by distinct RA-projecting neuron groups. From this assumption, the output sequence of RA is a higher-order Markov process, even though the RA-projecting neurons in the HVC fire on first-order Markov processes. We developed a neural network model of the local circuits in the HVC that explains the mechanism by which RA-projecting neurons transit stochastically on first-order Markov processes. Numerical simulation showed that this model can generate first-order Markov process song sequences.     

Imprinting and the origin of parasite-host species associations in brood-parasitic indigobirds, Vidua chalybeata

Brood-parasitic village indigobirds, Vidua chalybeata, were bred in captivity and foster-reared by their normal host species, the red-billed firefinch, Lagonosticta senegala, or by an experimental foster species, the Bengalese finch, Lonchura striata. Captive-reared female indigobirds were tested as adults for mate choice and for host choice. In tests of mate choice, female indigobirds responded preferentially towards mimicry songs of male indigobirds that were similar to those of the females' own foster parents. Females reared by Bengalese finches responded to male songs that mimicked Bengalese finch song rather than to male songs that mimicked their normal host species, the firefinch. In tests of host choice, females reared by Bengalese finches laid in the nests of Bengalese finches, and females reared by firefinches laid in the nests of firefinches. Wild-caught females showed the same behaviours as captive-bred females reared by firefinches. A female indigobird's social companions (firefinch or Bengalese) following her independence of her foster parents had no effect on her sexual response to male mimicry song or her choice of a host species in brood parasitism. The results support the predictions of a model of imprinting-like behaviour development in which young indigobirds focus their attention on their foster parents, rather than a model of innate bias for songs and nests of their normal host species, or a null model of nonspecific brood parasitism and differential survival. The results provide experimental support for the recent origin of brood parasite-host associations and the significance of imprinting in speciation in these brood parasites. Copyright 2000 The Association for the Study of Animal Behaviour.     

Role of a telencephalic nucleus in the delayed song learning of socially isolated zebra finches

Male zebra finches normally learn their song from adult models during a restricted period of juvenile development. If song models are not available then, juveniles develop an isolate song which can be modified in adulthood. In this report we investigate the features of juvenile experience that underly the timing of song learning. Juvenile males raised in soundproof chambers or in visual isolation from conspecifics developed stable isolate song. However, whereas visual isolate song notes were similar to those of colony-reared males, soundproof chamber isolates included many phonologically abnormal notes in their songs. Despite having stable isolate songs, both groups copied new notes from tutors presented to them in adulthood (2.7 notes per bird for soundproof chamber isolates, 4.4 notes per bird for visual isolates). Old notes were often modified or eliminated. We infer that social interactions with live tutors are normally important for closing the sensitive period for song learning. Lesions of a forebrain nucleus (IMAN) had previously been shown to disrupt juvenile song learning, but not maintenance of adult song for up to 5 weeks after surgery. In this study, colony-reared adult males given bilateral lesions of IMAN retained all their song notes for up to 4–7.5 months after lesioning. However, similar lesions blocked all song note acquisition in adulthood by both visual and soundproof chamber isolates. Other work has shown that intact hearing is necessary for the maintenance of adult zebra finch song. We infer that auditory pathways used for song maintenance and acquisition differ: IMAN is necessary for auditorily guided song acquisition—whether by juveniles or adults—but not for adult auditorily guided song maintenance. © 1993 John Wiley & Sons, Inc.     

Forebrain circuits underlying the social modulation of vocal communication signals

Across vertebrate species, signalers alter the structure of their communication signals based on the social context. For example, male Bengalese finches produce faster and more stereotyped songs when directing song to females (female‐directed [FD] song) than when singing in isolation (undirected [UD] song), and such changes have been found to increase the attractiveness of a male's song. Despite the importance of such social influences, little is known about the mechanisms underlying the social modulation of communication signals. To this end, we analyzed differences in immediate early gene (EGR‐1) expression when Bengalese finches produced FD or UD song. Relative to silent birds, EGR‐1 expression was elevated in birds producing either FD or UD song throughout vocal control circuitry, including the interface nucleus of the nidopallium (NIf), HVC, the robust nucleus of the arcopallium (RA), Area X, and the lateral magnocellular nucleus of the anterior nidopallium (LMAN). Moreover, EGR‐1 expression was higher in HVC, RA, Area X, and LMAN in males producing UD song than in males producing FD song, indicating that social context modulated EGR‐1 expression in these areas. However, EGR‐1 expression was not significantly different between males producing FD or UD song in NIf, the primary vocal motor input into HVC, suggesting that context‐dependent changes could arise de novo in HVC. The pattern of context‐dependent differences in EGR‐1 expression in the Bengalese finch was highly similar to that in the zebra finch and suggests that social context affects song structure by modulating activity throughout vocal control nuclei. © 2015 Wiley Periodicals, Inc. Develop Neurobiol 76: 47–63, 2016     

Adult bengalese finches (Lonchura striata var. domestica) require real-time auditory feedback to produce normal song syntax

Songbirds develop their songs by imitating songs of adults. For song learning to proceed normally, the bird's hearing must remain intact throughout the song development process. In many species, song learning takes place during one period early in life, and no more new song elements are learned thereafter. In these so-called close-ended learners, it has long been assumed that once song development is complete, audition is no longer necessary to maintain the motor patterns of full song. However, many of these close-ended learners maintain plasticity in overall song organization; the number and the sequence of song elements included in a song of an individual vary from one utterance to another, although no new song elements are added or lost in adulthood. It is conceivable that these species rely on continued auditory feedback to produce normal song syntax. The Bengalese finch is a close-ended learner that produces considerably variable songs as an adult. In the present study, we found that Bengalese finches require real-time auditory feedback for motor control even after song learning is complete; deafening adult finches resulted in development of abnormal song syntax in as little as 5 days. We also found that there was considerable individual variation in the degree of song deterioration after deafening. The neural mechanisms underlying adult song production in different species of songbirds may be more diverse than has been traditionally considered. © 1997 John Wiley & Sons, Inc. J Neurobiol 33: 343–356, 1997     

Social influences during song development in the song sparrow: a laboratory experiment simulating field conditions

Oscine songbirds are exposed to many more songs than they keep for their final song repertoire and little is known about how a bird selects the particular song(s) to sing as an adult. We simulated in the laboratory the key variables of the natural song learning environment and examined the song selection process in nine hand-reared male song sparrows, Melospiza melodia, a species in which males sing 5-11 song types. During their second and third months (their presumed sensitive period), subjects were rotated equally among four live adult male tutors that had been neighbours in the field. Tutors were housed in individual aviary 'territories' in four corners of the roof of a building; subjects could see only one tutor at a time, but they could hear the others at a short distance. Later in their first year (months 5-12), half the subjects were again rotated among all four tutors and the other half were randomly stationed next to just one tutor. Results from this experiment confirm and extend the findings from our two previous field studies of song learning in this species. Young males in this experiment (1) learned whole song types, (2) learned songs from multiple tutors, (3) preferentially learned songs that were shared among their tutors, (4) learned songs that other young males in their group also chose, and (5) learned more songs from the tutor they were stationed next to during the later stage (stationary subjects). These last two results support the late influence hypothesis that interactions after a bird's sensitive period affect song repertoire development. Copyright 2000 The Association for the Study of Animal Behaviour.     

Song learning in domesticated canaries in a restricted acoustic environment

Many songbirds learn their songs early in life from a song model. In the absence of such a model, they develop an improvised song that often lacks the species-typical song structure. Open-ended learners, such as the domesticated canary, are able to modify their songs in adulthood, although the mechanisms that guide and time the song-learning process are still not fully understood. In a previous study, we showed that male domesticated canaries lacking an adult song model in their first year substantially change their song repertoire and composition when exposed to normally reared conspecifics in their second year. Here, we investigate song development in descendants of canaries that were raised and kept as a peer group without a song model. Such males represent tutors with abnormal song characteristics. Interestingly, the F₁ generation developed quite normal song structure, and when brought into an environment with normally raised canaries in their second year, they did not modify their songs substantially. These results suggest that contact with an adult song model early in life is crucial for song crystallization, but also that song development is at least partly guided by innate rules. They also question the existing classification of canaries as open-ended learners.     

The Onset of Song Learning and Song Tutor Selection in Fledgling Zebra Finches

We investigated song development in the pre‐independent zebra finch (aged 15–35 d), a period when neural pathways for song learning and production are forming and social influences outside the family are limited. Expt 1 investigated the onset and the minimum duration of tutoring needed for song learning in fledglings. We found most begin to learn song from 25 d of age and need about 10 d contact with the father tutor to make accurate copies. This onset corresponds with major developments in the formation of the neural circuitry implicated in song acquisition. Subsong also begins on day 25 suggesting that the sensory and motor phases of song learning fully overlap in the zebra finch. Our findings support the hypothesis that the song circuitry is fully functional by 35 d of age and the sensitive phase for zebra finches extends therefore from about days 25–65. However, only the first 10 d of this period are necessary to learn a tutor's song with fair accuracy. Expt 2 investigated the role of the paternal bond, spatial proximity and mating status in a fledgling's choice of song tutor where the father was the sole parent. Young chose the father over single unrelated males (expt 2a) or unrelated males in company with their female partners (expt 2b). Given the close spatial proximity of both potential tutors to the fledglings it is probably the filial bond, established via paternal care that is the cause of this preference. Zebra finches sing the same song phrase in two contrasting contexts: female‐directed song during pre‐copulatory courtship and undirected song where no female or display is involved. In expt 3 we tested the song learning preference of pre‐independent young for two categories of non‐paternal tutors: those singing predominantly female‐directed song and those singing exclusively undirected song. There was a small, but significant, preference for fledgling zebra finches to copy songs from males that sang female‐directed song. This preference is consistent with the hypothesis that young males not only learn the acoustic features of their tutor's song but also the visual and dynamic movements that constitute the courtship display.     

Song learning and social interaction in indigo buntings

Indigo buntings (Passerina cyanea) isolated as individuals from 60 days of age developed abnormal songs. Birds isolated in groups for 9 to 10 months and then individually isolated developed slightly more normal songs but lacked the adult song figures. Birds copied the songs of adult tutors with whom they interacted socially. Birds with two tutors copied the songs of tutors that they could see and join in supplanting behaviour, but not songs of tutors from which they were visually isolated. One song was transmitted culturally across three generations under experimental conditions. The importance of social factors in song development of yearling buntings explains the development of local groups of males that share songs or dialects with each other in the field.     

An Automated Procedure for Evaluating Song Imitation

Songbirds have emerged as an excellent model system to understand the neural basis of vocal and motor learning. Like humans, songbirds learn to imitate the vocalizations of their parents or other conspecific “tutors.” Young songbirds learn by comparing their own vocalizations to the memory of their tutor song, slowly improving until over the course of several weeks they can achieve an excellent imitation of the tutor. Because of the slow progression of vocal learning, and the large amounts of singing generated, automated algorithms for quantifying vocal imitation have become increasingly important for studying the mechanisms underlying this process. However, methodologies for quantifying song imitation are complicated by the highly variable songs of either juvenile birds or those that learn poorly because of experimental manipulations. Here we present a method for the evaluation of song imitation that incorporates two innovations: First, an automated procedure for selecting pupil song segments, and, second, a new algorithm, implemented in Matlab, for computing both song acoustic and sequence similarity. We tested our procedure using zebra finch song and determined a set of acoustic features for which the algorithm optimally differentiates between similar and non-similar songs.     

Song development in the white-crowned sparrow: social factors and sex differences

Results of earlier studies indicated that hand-raised white-crowned sparrows exposed to taperecorded songs learned conspecific song between ages 10–50 days, but not before or after that age. These studies also indicated that allospecific songs were not learned. We describe song development in 41 male and 22 female hand-raised white-crowned sparrows. Thirty males and 15 females were exposed to a live adult singing male. It was found that most male students learned the song of their live tutor even though tutoring was begun at 50 days of age, an age by which young would have dispersed from the natal to the breeding area. Male students learned allospecific song as easily as they did conspecific song, even though conspecific song was present in the laboratory. Only three females copied any part of the song of either conspecific or allospecific live tutors. Six 50-day-old males and seven females were exposed to taperecorded song and none learned the tutor song. These results indicate that there are sex differences in song learning, and that, if live tutors are used, the sensitive phase for male song learning extends beyond 50 days of age. We conclude that social interaction can override any auditory gating mechanism that prevents inappropriate stimuli from influencing song learning centres.     

Hatching late in the season requires flexibility in the timing of song learning

Most songbirds learn their songs from adult tutors, who can be their father or other male conspecifics. However, the variables that control song learning in a natural social context are largely unknown. We investigated whether the time of hatching of male domesticated canaries has an impact on their song development and on the neuroendocrine parameters of the song control system. Average age difference between early- and late-hatched males was 50 days with a maximum of 90 days. Song activity of adult tutor males decreased significantly during the breeding season. While early-hatched males were exposed to tutor songs for on average the first 99 days, late-hatched peers heard adult song only during the first 48 days of life. Remarkably, although hatching late in the season negatively affected body condition, no differences between both groups of males were found in song characteristics either in autumn or in the following spring. Similarly, hatching date had no effect on song nucleus size and circulating testosterone levels. Our data suggest that late-hatched males must have undergone accelerated song development. Furthermore, the limited tutor song exposure did not affect adult song organization and song performance.     

Song Learning in Zebra Finches Exposed to a Series of Tutors during the Sensitive Phase

Young male zebra finches exposed to a series of adult males during their sensitive phase for song learning tended to learn from several of them, in contrast to those exposed to two or more tutors simultaneously. There is a tendency to learn more from later tutors, either because the memory of their elements over-writes that of earlier ones or because sensitivity is greater at this stage. No clear relation could be found between features of the song or behaviour of tutors and whether or not they were copied. In common with earlier studies, tutor song output, beyond a certain minimum was found to be unimportant.     

Complex sequencing rules of birdsong can be explained by simple hidden Markov processes

Complex sequencing rules observed in birdsongs provide an opportunity to investigate the neural mechanism for generating complex sequential behaviors. To relate the findings from studying birdsongs to other sequential behaviors such as human speech and musical performance, it is crucial to characterize the statistical properties of the sequencing rules in birdsongs. However, the properties of the sequencing rules in birdsongs have not yet been fully addressed. In this study, we investigate the statistical properties of the complex birdsong of the Bengalese finch (Lonchura striata var. domestica). Based on manual-annotated syllable labeles, we first show that there are significant higher-order context dependencies in Bengalese finch songs, that is, which syllable appears next depends on more than one previous syllable. We then analyze acoustic features of the song and show that higher-order context dependencies can be explained using first-order hidden state transition dynamics with redundant hidden states. This model corresponds to hidden Markov models (HMMs), well known statistical models with a large range of application for time series modeling. The song annotation with these models with first-order hidden state dynamics agreed well with manual annotation, the score was comparable to that of a second-order HMM, and surpassed the zeroth-order model (the Gaussian mixture model; GMM), which does not use context information. Our results imply that the hierarchical representation with hidden state dynamics may underlie the neural implementation for generating complex behavioral sequences with higher-order dependencies.