Predicting the Learning Diffusion and Inheritance of Behavioural Innovations in Animal Populations.

The aim of this research is to increase our understanding of the processes by which new knowledge and skills is transferred between animals, including humans, through copying ('social learning'). For human beings, imitation and teaching are thought to be the most important processes by which knowledge spreads. However, in other animals there is little evidence that teaching occurs but nonetheless inexperienced and young individuals pick up clues from more knowledgeable others concerning, for instance, the identity of predators or how best to process foods. This social learning can generate 'traditions' for performing particular behaviour patterns, for instance, eating specific foods, or taking particular pathways. As a consequence, the behaviour of animals may vary from one population to the next, not just because they possess different genes, or are exposed to different environmental resources, but also because they have learned different habits from more experienced members of their population. A challenge for researchers studying animal behaviour is to identify these 'traditions' and to work out how novel behaviour and skills ('innovations') spread through populations. This research programme will use mathematics and statistics to try to determine when animals acquire their behaviour through social learning, and how novel behaviours spread through social learning in animal populations. These mathematical predictions will be tested with laboratory experiments with birds and through analyses of chimpanzee and dolphin behaviour. The information generated will be useful to a variety of researchers, including biologists, psychologists and economists.

Interest in social learning, the process whereby animals acquire skills and knowledge by observing other animals, has also been fuelled by high-profile reports of inter- and intra-population variation in the behavioural repertoires of apes, monkeys and cetaceans, and of the rapid spread of novel behavioural innovations in vertebrate populations. However, claims that such variation represent animal 'cultures' remain controversial in the absence of clear methods for ruling out alternatives, such as genetic differences between populations, or asocial learning in response to varying environmental resources, while field researchers have generally been unable to prove that innovations spread by social learning. New methods are required allowing researchers to (1) identify and quantify social learning among animals living in groups, (2) explain and predict the rate, path and pattern of diffusion of learned innovations, and (3) quantify the extent to which variance in the inheritance of acquired behavioural traits is attributable to social learning, as opposed to other sources. This project describes and develops novel mathematical and statistical methods that have the potential to address these goals. They include the bootstrapping of probability distributions for the asocial learning of target behaviour, the identification of statistical 'signatures' of cultural transmission, and the development of regression models for the inheritance of complex behavioural traits. The methods will be developed using experimental data derived from a captive budgerigar population and tested on chimpanzee and dolphin data. The experiments and theory together will lead to a deeper understanding of social learning in animals, including humans, and will be of broad interest to behavioural biologists, psychologists, economists and to industry.