On the significance of song amplitude in birds – function, mechanisms, and ontogeny

Abstract

Birdsong is an important model in the study of animal behaviour, environmental adaptation and sexual selection. Its main functions are territory defense and attraction of sexual partners. Birdsong is a multidimensional trait. Each of its dimensions may be subject to different selection pressures, may be constrained by different costs and may signal different qualities. For example, a high song rate may be important to defend a territory, may be constrained by time budgets, and may signal a high foraging efficiency to females. A large song repertoire may be important to attract mates, be subject to developmental constraints and may signal early nutrition.One song trait that so far has received little attention is amplitude, i.e. the sound pressure level of a song. It has been shown that there is a considerable variation in song amplitude between males, but the causes and consequences of this variation are poorly studied. In my dissertation, I performed several experiments in order to shed some light on the function and evolution of song amplitude in birds.In a first step, I asked whether song amplitude is subject to sexual selection through female choice and male-male competition. To this end, I tested song preferences of female zebra finches and I simulated high-amplitude and low-amplitude intruders in territories of male chaffinches. My results demonstrated that female zebra finches have a preference for loud songs over soft songs if all other song parameters are kept constant, and that male chaffinches react considerably stronger to loud territory intruders than to intruders singing with low amplitude. This suggests that song amplitude is likely to be sexually selected. Moreover, the strong territorial response of chaffinches towards loud intruders provides the first experimental evidence that song amplitude may be constrained by social aggression. But why should females prefer loud males and why should loud males elicit stronger territorial aggression?To answer these questions, I investigated whether song amplitude may have a signalling value. In particular, I tested whether song amplitude may reflect testosterone level or body condition, and thus signal motivation or male quality. To this end, I manipulated testosterone levels of adult male Bengalese finches and the nutritional status of adult male zebra finches. I found that elevated testosterone levels did not affect song amplitude in Bengalese finches, but that zebra finches with artificially decreased body weight sang significantly less loud than control birds. This demonstrates for the first time that song amplitude may signal the current condition of the singer and suggests that females may make use of this information when choosing a mate.Finally, I studied the effects of vocal learning on song amplitude. I found that young zebra finches adopted the amplitude of song elements from their tutors, which suggests that amplitude is either learned, bound to spectro-temporal element properties due to production constraints, or both. My findings indicate that vocal production learning may guide the ontogenetic trajectory of vocal amplitude towards a default level which is then modified in the short-term by the current condition of the adult male.My findings contribute to a deeper understanding of the causes and consequences of song amplitude in birds and highlight the importance of signal intensity in acoustic communication. Moreover, they emphasize that behaviour should be viewed as a complex interplay between mechanistic, functional and ontogenetic aspects.