Harnessing learning biases is essential for applying social learning in conservation

Social learning can influence how animals respond to anthropogenic changes in the environment, determining whether animals survive novel threats and exploit novel resources or produce maladaptive behaviour and contribute to human-wildlife conflict. Predicting where social learning will occur and manipulating its use are, therefore, important in conservation, but doing so is not straightforward. Learning is an inherently biased process that has been shaped by natural selection to prioritize important information and facilitate its efficient uptake. In this regard, social learning is no different from other learning processes because it too is shaped by perceptual filters, attentional biases and learning constraints that can differ between habitats, species, individuals and contexts. The biases that constrain social learning are not understood well enough to accurately predict whether or not social learning will occur in many situations, which limits the effective use of social learning in conservation practice. Nevertheless, we argue that by tapping into the biases that guide the social transmission of information, the conservation applications of social learning could be improved. We explore the conservation areas where social learning is highly relevant and link them to biases in the cues and contexts that shape social information use. The resulting synthesis highlights many promising areas for collaboration between the fields and stresses the importance of systematic reviews of the evidence surrounding social learning practices.