The functional role and female perception of male song in Zebra Finches

The song of male Zebra Finches has been the focus of decades of behavioural, developmental, neurobiological and, increasingly, genomic research. Zann was the first to summarise the immense and integrative research effort in a landmark synthesis of field and laboratory studies of Zebra Finches, which paralleled his own championing work on the sociality and vocal behaviour of estrildid finches in the wild and in captivity. The study of the production and perception of Zebra Finch song has driven theoretical, empirical and technological advances in behavioural ecology, endocrinology and neuroethology, and led to a greater understanding of the evolution of animal communication systems in general. A survey of the literature shows that there are still significant gaps in our understanding of the mechanisms underlying the responses of non-singing females to male sexual displays, including song. We focus on recent insights into the features and functions ofmale song that shape female choice, regardingbothbehavioural andneurobiologicalmeasures of selectivity. This review underscores the need for continued research into the biological mechanisms underlying the perception of male song by female Zebra Finches and confirms this system as a valuable and productive model for research on animal communication. Additional keywords: Bengalese Finch, female choice, genome, Lonchura striata vars. domestica, neuroethology, sexual selection, Taeniopygia guttata. Structure and development of Zebra Finch song The song of the Zebra Finch (Taeniopygia guttata), produced solely by males, is a short (1–2 s) vocalisation comprised of harmonic stacks and broadband notes, called syllables, that are produced in stereotyped sequences calledmotifs. Songmotifs are repeated in bouts (Sossinka and Bohner 1980). The number, acoustic features and diversity of syllables that are produced are stereotyped within individuals but vary between individuals (Immelmann 1969). Individual differences allow the songs of male Zebra Finches to be easily distinguished by the human listener and also enable true individual recognition of mates (Miller 1979a) and fathers (Miller 1979b; Slater et al. 1988) by females (see below). Although the female Zebra Finch typically does not sing, masculinisation via oestradiol administration during ontogeny can induce song production, indicating an important role for gonadal hormones in the organisation of sexual dimorphism of vocal displays (Adkins-Regan et al.1994).Genetic determination of sex (Agate et al. 2003), coupled with neuroendocrine effects up to the first week after hatching, are most influential for this development of the species-typical sexual differentiation in singing and the neuroanatomical basis of male-only song (Adkins-Regan et al. 1994). However, the administration of testosterone to adult females can still induce and increase the frequency of female singing (Adkins-Regan 1999). In parallel, ontogenetic oestradiol treatment causes female Zebra Finches to sing in adulthood, and shows that these songs are learned during development, in parallel to the development of males’ imitative learning of song displays (Simpson and Vicario 1991). The male Zebra Finch copies his song from an adult tutor during well-defined developmental stages (Slater et al. 1988; Zann 1996). Song ontogeny comprises memorising a song template, followed by motor production of subsong and plastic song, until the song becomes crystallised at ~100 days of age. Following the crystallisation, typically only minor adjustments are made in the acoustic characteristics of the song. Such plasticity can include some transient deviations in song pattern, including syllable deletions or alterations, that are likely to reflect changes in the neural motor control pathway in the forebrain (Helekar et al. 2000), or amplitude modulation in response to the distance of the intended recipient of the vocal display (Brumm and Slater 2006). But the likelihood and functional relevance of these anatomically and socially induced aspects of acoustic plasticity remain untested in the wild. In parallel with males, female Zebra Finches learn to remember the song of their fathers and familiar brothers during development in captivity (Riebel et al. 2002; Riebel 2003), and also learn to recognise the songs of their mates during adulthood (Miller 1979a) but, again, the function of familiarity-based discrimination of songs by females in nature remains unclear. Zebra Finches of both sexes are able to learn about conspecific song from playbacks of vocalisations (Bolhuis et al. 1999), although males will more thoroughly and accurately imitate songs of tutors with whom they can vocally and visually CSIRO PUBLISHING Review www.publish.csiro.au/journals/emu Emu, 2010, 110, 209–218 Royal Australasian Ornithologists Union 2010 10.1071/MU10003 0158-4197/10/030209 interact (Eales 1989). Evidence from blindfolded juvenile males has further shown that physical interaction itself, even in the absence of vision, enhances the learning of song (Adret 2004). This is in contrast to the social inhibition of song imitation by several brothers in the same cage (Tchernichovski and Nottebohm 1998) and presents a strong case for further investigations into the role that the sex and numbers of siblings might play during development in both male imitation of, and female preference for, tutor songs. In the absence of a live song tutor, a male Zebra Finch will develop a rudimentary version of the adult species-typical song that is characterised by abnormally high frequencies and long syllables, often repeated in bouts that lack the normal stereotypy in syllable order of adults (Williams et al.1993). This type of song is referred to as isolate song. Previously untutored adult Zebra Finches have also been shown to copy new syllables from tutors or exemplars to which they are exposed at an age that is beyond the typical sensitive period of song learning in this species (Eales 1985; Morrison and Nottebohm 1993). Imitation and improvisation occur in each generation of males to produce normal (social) adult song. A recent experiment tracked the change in features of male song across multiple generations, but started with males that sang isolate songs as tutors (Fehér et al. 2009). The results were unexpected in that normal, species-typical song structure re-established themselves within approximately four generations in a closed microcosm where males of the previous generation served as song tutors for the next generation of males. Such novel experimental paradigms and surprising findings also have critical implications and require new theory to explain the evolutionary benefit and cost of investment into the neural substrates of imprinting, copying and learning to generate song cultures. For example, pinpointing the relative roles of species identity versus tutoring experience in shaping generational cultural shifts could involve documenting the process of cultural evolution of song, and the impacts on female perception, across generations, starting with male Zebra Finch tutors singing heterospecific songs (e.g. through cross-fostering with Bengalese Munias (or Bengalese Finches, Lonchura striata vars. domestica) as parents; Clayton 1989; Campbell and Hauber 2009a). Using cross-fostered males as initial tutors would be especially valuable because male Zebra Finches are imperfect mimics of Bengalese Munia songs: whereas they copy syllable structure closely, they retain some syntax and rhythm of Zebra Finch songs (e.g. Woolley et al. 2010). In turn, the neurophysiological basis of preferences in female Zebra Finches between normal songs and Bengalese Munia-like songs of conspecific males is still unresolved (Campbell and Hauber 2010b). Such new manipulations would thus both inform and challenge traditional hypotheses for the roles and consequences of early experience and learning in shaping species specificity in female choice for conspecific songs (see below). Song and female mate-choice Male song is one of the important cues for mate-choice in female Zebra Finches. Field and laboratory observations indicate that males readily approach single females, whereas females are more discriminating in their spatial approaches (Zann 1996; Campbell and Hauber 2009b). This is in concordance with general predictions of intersexual differences in the thresholds of mate-discrimination, as females bear a greater reproductive cost of erroneous mate-choice than do males, even in biparental species (Hauber and Lacey 2005). Choosing a suitable social partner and, eventually, a mate is essential in Zebra Finches in general, and for females in particular, because the pair-bond is lasting; separation apparently occurs only with the death of one partner (Zann 1996), and new research has confirmed that extra-pair fertilisations are rare in natural populations (Birkhead 2010; Griffith et al. 2010). Additionally, in the wild, the probability of offspring survival to sexual maturity is estimated to be as low as one in four (Zann 1996) and so a female’s choice of a mate, who is to provide parental care and genetic contributions to her offspring, has a critical impact on her own reproductive success, through both the physical condition and the genetic quality of the male. Zebra Finches are not territorial and males sing to females at close range,within loosebreeding colonies and largeflocks (Zann 1996). Song is thus typically accompanied by other sensory and social cues, including colour of plumage and bill, body size and courtship dance. Severalmalemorphological traits serve as visual cues for female Zebra Finches, signalling information that is relevant to mate-choice. Immelmann (1959), working on captive birds, initially identified bill-colour, eye-stripes, and tail-bands in male Zebra Finches as visual conspecific recognition signals for females. Since then, studies that tapped into the natural variation of vocal and visual sexual cues, and the experimental alteration of these traits, repeatedly demonstrated that hearing male vocalisations and perceiving other sensory cues, including colour of plumage and bill, interact to influence female attraction to potential mates (Campbell and Hauber 2009c, 2010b). The relative salience and role of these cues in female choice within natural populations, however, remain unclear, as does their developmental trajectory, with Richard Zann’s own last study on wild-caught birds demonstrating opposing effects of nutritional constraint during development in reducing the complexity of song displays, but not the accuracy of song copying, of male Zebra Finches (Zann and Cash 2008). Although multimodal sensory cues are clearly involved in female mate-choice, including the synergistic effects of vision and hearing (Campbell and Hauber 2010a), many laboratory studies have shown that song alone can direct the choice behaviour of a female and predict which males are attractive (Miller 1979a, 1979b; Clayton 1990;Riebel et al.2002; Tomaszycki and Adkins-Regan 2005; Holveck and Riebel 2007; Vignal et al. 2008; Campbell and Hauber 2009a; Vyas et al.2009).Next,wediscuss the roleofmale song in femalemateselection and how song preferences of females are related to pairing, mating and reproductive behaviours. Female behaviours that predict mate-preference To understand the role of male song in mate-choice by females, it is critical to describe and quantify female displays of social discrimination, sexual preferences, male attraction and mating decisions. Mate-choice-related behaviours in female Zebra Finches can be broadly grouped into three major classes. First, 210 Emu M. E. Hauber et al.