The evolutionary roots of vocal learning: Exploring vocal learning abilities in vocal non-learners in birds

Abstract

Vocal learning is a complex behaviour that enables animals to imitate and learn sounds from their environment. Many aspects of this behaviour remain poorly understood, particularly how prevalent it is, how it evolved and whether it is a binary behaviour (present or absent) or expresses along a continuum. One way to explore the evolutionary roots of vocal learning is to focus on its precursors like rudimentary vocal learning or innate vocal behaviour. Birds are an excellent study model to understand the evolution of vocal learning because three groups of birds are vocal learners in birds: parrots, hummingbirds, and songbirds (oscines). In birds, there is also compelling evidence that suboscines that share a common ancestor with parrots and oscines have rudimentary vocal learning abilities. To better understand how vocal learning evolved, I conducted a comparative study of an innate call, the distress call, among avian species with different vocal learning abilities and showed that the evolution of innate calls in birds follows a continuum between vocal non-learners and birds that learn their songs. I also showed that innate calls evolved faster in song learners (songbirds and hummingbirds) compared to non-song learners such as parrots or vocal non-learners. These results have broad implications for understanding the evolutionary diversification processes of innate vocalizations which reveal patterns of correlated evolutionary changes, best known as evolutionary integration, with vocal learning. Next, I explored the vocal learning abilities of the New Zealand wrens (Acanthisitti), a critical group of birds that shares a common ancestor with vocal learners (parrots and oscines) and suboscines. New Zealand wrens have been characterised as vocal non-learners; however, this has never been explicitly tested. I demonstrated that one of the two extant species of New Zealand wrens, the titipounamu (rifleman, Acanthisitta choris) has individual and group vocal signatures at nests, but no vocal matching (i.e. exact copy of sound) at the nestling stage. I also showed that sociallyclose unrelated titipounamu share high acoustic similarities in their call structure which are strongly shaped by social interactions. These findings suggest that titipounamu can vocally converge – a form of vocal imitation. Overall, this thesis shows that vocal convergence was likely present in the common ancestors of the New Zealand wrens, parrots and songbirds. This research contributes a number of insights into the evolution of vocal learning in birds and provides a basis for future research interested to explore vocal learning abilities in vocal non-learners.