Résumé
The overwhelming focus of studies on communication has been on interactions among conspecifics. However, communication is often selected by a complex network of disparate intended perceivers and eavesdroppers belonging to multiple species. Shifting towards a multiple-perceiver paradigm requires a framework to compare the perception of signals across the different community members. Here, we present the stimulation landscape, a generalist model that achieves this goal. A stimulation landscape consists of a multidimensional space describing every possible stimulus for a given signalling system, to which is added one dimension indicating for each stimulus its conspicuousness for a given perceiver. Random sampling of stimuli in the landscape then allows computing a reference distribution of conspicuousness, which is used to standardize the observed conspicuousness. The stimulation landscapes corresponding to the different perceivers participating in a communication system all have the same dimensionality and scaling. They can thus be combined to describe the amalgamation of the selective forces exerted on a signal by the distinct sensory systems of multiple perceivers. We detail the model in the context of colour signalling and apply it to the case of sexual communication in songbirds. Songbirds have a different visual system than their main predators, the birds of prey. We therefore asked whether songbirds evolved colour signals that can mediate the trade-off between sexual selection exerted by conspecifics and natural selection exerted by birds of prey. We show that yellow not ultraviolet colours as previously thought maximize the difference of conspicuousness to songbirds and to birds of prey, but that the perceptual similarities between these two groups generally hamper the evolution of private visual communication in songbirds. The stimulation landscape is a valuable tool to investigate the role of communication in structuring the large networks of interactions between species. We further show that the stimulation landscape is related to the evolutionary model of adaptive landscape. By explicitly addressing communication in a community context, the stimulation landscape contributes to bridge the current gap between evolution and ecology.