In 1963, a landmark paper by Niko Tinbergen laid out the aims and methods of ethology and, in so doing, extended and clarified Julian Huxley's classification of the different ways in which one can investigate biological processes. I discuss the status of one of these "four Why questions", that of function or survival value, and the relationship of Tinbergen's ethology to behavioural ecology, the main field asking functional questions about animal behaviour today. Function itself can be defined in many different ways and behavioural ecologists themselves use it both in the context of current utility and selective history. I review these definitions in the light of analyses by philosophers of science, behavioural ecologists and, of course, Tinbergen's own use of the word. I defend the view, accepted by many philosophers of science, that the definition of 'function' must have a historical component, both to avoid teleology and to retain the everyday sense of questioning 'What is it for?' That said, in reviewing the different methods that can be used to determine function, I defend the view that investigations of current utility, as practised by behavioural ecologists, can provide the most important clues to the selective forces that have shaped behaviours. Finally, I consider the evolution of the discipline of behavioural ecology, its current status and future prospects.
Recent theoretical and experimental studies appear to provide a comprehensive explanation for the evolutionary, and short-term, stability of biparental care in birds. However, much of the intraspecific variation in the absolute and relative contribution by the male and female remains to be explained. Most studies of the natural variation across pairs reveal positive correlations between the level of male and female nest defence or brood provisioning, but some species show negative relations, or between-season variability in the direction of the relationship. This study examines the determinants of male and female provisioning rate in monogamous European starlings (Sturnus vulgaris), in terms of the patterns of male display, pairing patterns, laying phenology and clutch attributes. There was assortative pairing by body size and, controlling for female body size, larger males attracted relatively fecund mates. Males that sang relatively more, controlling for these body size effects, were paired with females that laid earlier. Female nest provisioning rates to experimentally standardized broods were positively correlated with female size, but male provisioning rates were unrelated to either song or any other attributes of themselves or their mates.
The functions of bird song are well described, but empirical studies examining the costs of singing are scarce. Two potential costs are a metabolic cost of singing, and lost feeding opportunities, but such energetic costs will only be biologically important if they have a significant effect on the bird's body reserves. Overnight loss of reserves has previously been found to increase with increasing song rates in nocturnally singing common nightingales Luscinia megarhynchos. However, it is not clear how such costs compare with those incurred by daytime-singing birds, which may forfeit foraging opportunities when they sing. In this paper we investigated the effect of variation in song rate on the body reserves of a typical daytimesinging bird, the European robin Erithacus rubecula, singing at different rates in three different circumstances: (i) Natural variation in song rate of free-living robins. (ii) Manipulations of the song rate of free-living robins using playbacks of conspecific song. (iii) Manipulations of the song rate of aviary-housed robins using playbacks of conspecific song. In all three parts of our study, birds gained less mass when they sang more. Our analyses suggested that this might have been due primarily to a reduction in food intake rate, rather than to the metabolic cost of the singing itself or a concurrent increase in locomotor costs. These results together demonstrate that the costs of singing, as measured by their overall net effects on body reserves, can have a significant impact on the energetic state of daytime-singing birds.