Owls show an astonishing variation in their degree of colour polymorphism, although the exact mechanisms driving such variation remain controversial. Here we address this fundamental question by considering information on all extant owls and recent advances in comparative methods in the frame of three mutually non-exclusive evolutionary scenarios. In addition, we study for the first time whether the evolution of influential ecological characters facilitated the evolution of colour polymorphism (or vice versa). In agreement with the niche divergence hypothesis, we found that species living under more variable luminal conditions, i.e., species with diurnal and crepuscular habits and those inhabiting in a mixture of open and closed habitats, were more likely to show colour polymorphism. Correlated evolution analyses revealed that a change in the luminal niche might be a fundamental requisite for the evolution of colour polymorphism. Moreover, polymorphism was more frequent among owl species occupying lower trophic levels, which could be explained by a particularly high selection for crypsis on small predator owls. Our results, thus, provide support for the idea that colour polymorphism in owls is an adaptive character likely maintained by the selective advantage of morphs under different environmental conditions via disruptive selection mechanisms.

Birds, due to their multiple colourful displays, constitute a classic paradigm for the study of colour evolution. Although avian eyes are remarkably coloured, the functional basis behind inter‐specific variability in iris colouration remains poorly understood. Owls are an ideal system to shed light on the role of ecology in promoting iris colour evolution as they show inter‐specific variation in iris colour and in niche specialization with some species being strictly nocturnal and others active during the day. Owls perching for hunting at night might be unnoticed by both predators and their prey if they had dark irises, which would predict that dark irises were more likely to evolve in strictly nocturnal species than in diurnal ones. Using phylogenetic comparative models, we tested the camouflage hypothesis for eye colour. The proportion of dark‐eyed owl species is higher among strictly nocturnal owls than among diurnal ones. Ancestral state reconstruction revealed that the owl ancestor of the family Strigidae was more likely bright‐irided whereas the ancestor of the family Tytonidae was more likely dark‐irided. Our results show robust support for the coevolution of iris coloration and nocturnality in the owls, and suggest that shifting to a nocturnal niche would be a prerequisite leading to the evolution of dark eyes in owls. The specific evolutionary pathway by which iris coloration and activity rhythm coevolve, however, remains to be investigated further as we have found only partial support for the idea that dark irises in birds might be an adaptive feature evolved due to the selective advantage of concealment from undesired visual receptors.

Aim Ecogeographical rules link animal colours, especially those produced by melanin pigments, with variation in environmental conditions over wide geographical scales. In particular, Gloger’s rule, coined for endothermic animals in two versions, suggests that tegument darkness should increase at high temperature, as well as in highly humid environments. On the other hand, the thermal melanism hypothesis predicts that darker colourations should be more frequent in colder areas given their thermoregulation benefits. Location Global. Time period Contemporary. Major taxa studied Strigiformes. Methods Here, we provide a global comparative test of these contrasting expectations in all extant owls (n = 198 species), a group of nocturnal birds displaying huge variability in the degree of melanin‐based plumage colouration and environmental specialization. Combining analyses at both species and assemblage level, we assessed the climatic and environmental variables explaining variation in plumage lightness and redness across broad geographical gradients. Results Darker and redder owl phenotypes are more likely found near the equator. Species and assemblage level analyses reveal that owls have more frequently darker and redder plumages in warmer regions. In addition, owl species living in more vegetated areas are darker, and owl assemblages show darker colours in wetter areas. Main conclusions Global patterns of colour variation in owls do not fit expectations from the thermal melanism hypothesis but support Gloger’s rule. Our findings also stress that several alternative selective forces may explain climatic effects on colouration over large geographical scales. Experimental work is urged to uncover the possible mechanisms behind the detected associations between owl colour and environmental variables.