The Role of the Social Environment in the Development of a Male's Sexual Signal

Elaborate male sexual signals are well known, but their development is poorly understood. This is especially true for signals that are not fixed early in life, but instead develop and change over time according to environmental factors. I will ask how the social environment affects how males acquire and exaggerate flexible sexual signals. First, which individuals in the population have the greatest influence on the development of a male's signal? Influential individuals may be those that respond to a male's signal, e.g. other males in the population that subsequently alter their antagonism towards the signaler, or females that subsequently change their sexual interest in the signaler. Alternatively other signalers may be influential and the focal male can observe their signals and modify his own signal. These may include close neighbours, leading to regional 'cultures' of sexual signal, or males that are themselves sexually successful and exhibit attractive signals. Second, which males are most influenced by others in the populations and so most flexible in their signal? Younger, less experienced males may be likely to incorporate novel elements into their signals. Alternatively, less successful males, may try to develop a different, more successful signal. Finally, already successful males may further exaggerate their attractive signals and so maintain their disproportionate share of matings. Male bowerbirds produce extremely elaborate sexual signals. They build structures (bowers) and decorate them with large numbers of objects. Bowers and their decorations change over time, so that a male's signal is flexible both within and between years. Males use novel objects as decorations on their bowers, but are highly selective as to which objects are placed on the bower and where they are located. Females pay close attention to bowers, so that males with large numbers of certain objects gain higher levels of mating success. Males pay close attention to objects on neighbour's bowers, stealing them to use on their own bowers, or destroying the bowers if they contain excessive numbers of particular attractive objects. The readiness of males to incorporate novel objects, that are discrete and tangible, into their signal makes them suitable for my studies because it is possible to experimentally i) manipulate what objects appear on the bowers of different males, and ii) measure individual's preference (or rejection) to test objects. I will exploit long term observational data, and conduct three field experiments to test these hypotheses. The experiments will add novel objects to the bowers of different sets of males (own, neighbours, successful males) and test how preferences for the objects changes after exposure. Variance between the three experiments will reveal which individuals influence signal development. Variance between males within each experiment will reveal which individuals are most influenced. My previous work on bowerbirds has revealed that: males differ in their object use, and correspondingly differ in their mating success; different populations of bowerbirds differ in their object use; within a population, local clusters of males use similar suites of objects as decorations; males destroy the bowers of neighbours who display large numbers of especially attractive decorations; and males appear to 'learn' about bowerbuilding from other males. These are all 'snapshot' observations of specific behaviours at single points in time, each suggestive of how sexual signals develop. My proposed work extends this to consider explicit social processes that underlie and explain signal development. This work is important for understanding how elaborate sexual signals arise, vary and exaggerate, by considering the social factors that influence their development. Such work has wide application, explaining the development of other, much more common, flexible sexual signals such as song, movement and object use.

It is now accepted that, contrary to traditional models of sexual selection, sexual signals can be determined by environmental as well as genetic factors. The social environment in which a male's sexual signal develops, has received relatively little attention. Sexual signals, such as song, object use or complex behavioural movements, appear to differ in how they develop, according to the behaviour of other individuals with whom the signaling male interacts. Male bowerbirds exhibit signals comprising objects on a bower, and these can be experimentally manipulated, permitting work to be carried out that would be impossible on signals such as song. I propose three field experiments, accompanied by analysis of a long term data set, to test explicit hypotheses as to: 1) which other individuals in a population have the strongest effect on the development of a male's signal, either through their response to his signal, or because of the form of signal that they themselves are exhibiting; 2) which males are most likely to vary the development of their signal in response to the behaviour of other individuals, specifically their age, experience or sexual success. Experiments will involve adding novel objects to the bowers of different sets of males (own, neighbours, successful males), associating such presentations with positive or negative experiences, testing how preferences for the objects changes after exposure, using a series of presentation choice tests, and understanding causes of variance in preferences between treatments (addressing question 1) and between individual males (question 2). I will use a population of spotted bowerbirds that I initiated work on in 1998, and studied extensively subsequently. This work is important for understanding how non-genetic, social factors impact on the development of elaborate sexual signals, and explains why signals under strong selection do not reach fixation through variation maintained by a changing social environment.