Sexual signals and preferences play a central role in mate choice, sexual selection, and reproductive isolation. However, when such traits are learned rather than genetically fixed, maintaining their accuracy and alignment becomes a major challenge. Learning introduces variation across individuals that can accumulate into mismatches between signallers and receivers, raising the question of how learned mating traits remain sufficiently stable to function in communication and reproduction. Birdsong is one such trait. In songbirds, songs are acquired through vocal learning, and successful song development depends on young individuals attending to, discriminating, and learning from appropriate acoustic models. Understanding how early song responses arise, what shapes tutor related biases, and whether adult song reliably reflects broader aspects of male quality is therefore important for explaining the development and evolution of learned sexual signals.
In this thesis, I investigate these questions in the pied flycatcher (Ficedula hypoleuca), focusing on early song discrimination and the information content of adult song. In Chapter I, I examine whether nestlings discriminate between their social father's song and that of an unfamiliar local male. I show that 13-day-old nestlings respond more strongly to their social father's song, indicating that nestlings can distinguish socially familiar from unfamiliar conspecific songs at an early age.
In Chapter II, I test whether nestlings respond differently to simplified short songs and longer complex songs played back from either their social father or an unfamiliar local male. Since males sing shorter songs after pairing, nestlings are typically exposed to simplified rather than complex paternal song, raising the question of how song length shapes early responsiveness. Nestlings showed stronger postural begging responses to shorter songs regardless of familiarity, suggesting that both early auditory experience and intrinsic acoustic salience contribute to early song responsiveness.
In Chapter III, I investigate whether early discrimination of paternal song reflects inherited predispositions or early auditory experience using an embryonic cross-fostering experiment in the wild. Nestlings reared by either genetic or foster parents were exposed to songs of their social father, genetic father, and an unfamiliar local male, but showed no differential begging responses to the three treatments. Because begging was strongly influenced by nestling condition, these results suggest that behavioural assays may fail to reveal auditory discrimination when nestling motivation is low, and the data remain inconclusive about the relative contributions of experience and predisposition.
In Chapter IV, I shift to adult sexual signalling and test whether song complexity predicts cognitive ability in breeding males, using a novel foraging task and a detour reaching task. Contrary to the hypothesis that preferred song traits signal cognitive quality, I find no consistent positive relationship between song complexity and cognitive performance. The traits most strongly linked to female preference were either unrelated or negatively related to task performance, suggesting that song complexity is unlikely to serve as a straightforward indicator of cognition.
Together, these findings shed new light on how learned sexual signals emerge, are shaped during development, and are maintained in natural populations.