Relative effectiveness of blue and orange warning colours in the contexts of innate avoidance, learning and generalization

Warning colours can deter predators from attacking unpalatable or toxic prey. These colours are avoided by predators, either innately or by learned response through trial and error. Predators that have learned these signals can also generalize their responses to similar colours. Much of the previous research on warning colours has examined long-wavelength colours (e.g. orange and red) and little is known about short-wavelength coloration (e.g. blue) in warning signals, despite its presence in unpalatable animals in many different taxa. We determined whether prey with only blue coloration, only orange coloration, and with both colours elicited innate avoidance from a naïve predator (Gambel's quail chicks, Callipepla gambelii), as well as how effectively the signals were learned and generalized. Prey were composed of pieces of mealworm (palatable or unpalatable) beneath a coloured paper tent. Blue did not elicit any innate avoidance, whereas orange and a combination of both colours (orange-and-blue) were effective at deterring naïve predators. Predators were able to learn the association between colour and unpalatability most effectively for orange prey, but also learned this association for blue and for orange-and-blue prey. Predators showed asymmetrical generalization, in which birds were more likely to generalize from the colour they learned as unpalatable (conditioned stimulus) to orange prey and orange-and-blue prey than to blue prey. We conclude that orange is likely a more effective warning colour to these birds, but blue can still be learned alone and when displayed adjacent to orange.