Skip to main content

The effects of ecology and behaviour on the evolution of colouration in Coraciiformes


Babarovic, Frane et al. (2023), The effects of ecology and behaviour on the evolution of colouration in Coraciiformes, Dryad, Dataset,


What drives the evolution of plumage colour in birds? Bird colour is likely to be under both natural and sexual selection where natural selection may favour evolution towards crypsis or camouflage whereas sexual selection may favour evolution towards conspicuousness. The responses to selection are predicted to relate to species’ ecology, behaviour, and life history. Key hypotheses have focused on habitat and light environment, breeding strategy, territoriality, and hunting behaviour. We tested these potential causes of colour variation in the Coraciiformes, a colourful clade of non-passerine birds, using phylogenetic comparative methods and data on plumage colouration and brightness measured from museum specimens. We found that correlates of colour evolution in Coraciiformes vary across body regions and depend on the focal colour property (hue or brightness). While light environment showed widespread effects on colouration in multiple body regions for both hue and brightness, selection pressures related to behavioural characteristics had more spatially localized effects (e.g. territoriality on wing feather brightness and hunting strategy on belly hue). Our results reveal both general patterns that may hold across other bird clades and more nuanced effects of selection that are likely to be mediated through the visual ecology of the signaller and receiver and the behavioural characteristics of Coraciiform species.


Two sets of data were collected for this research.

First, in the Natural History Museum at Tring bird collection, we used bird skins of 135 Coraciiformes species to measure plumage colouration. Calibrated digital images of study skins were taken using methods described in Cooney et al. (2019). They were used to quantify both chromatic (hue and saturation) and achromatic (brightness) components of the colour. We draw a series of polygons on every image representing body patches. We covered 11 body patches on each specimen. We extracted RGB values from each polygon and then converted them to avian colourspace values (cone catch values: u, s, m, and l). These values represent the chromatic measurement of bird colour. We also extracted an achromatic measurement of colour as the stimulation value of double cones. Following, we have used Principal component analysis to reduce the dimensionality of cone catch values from 4 (i.e. u, s, m, l) to 2 values (PC1 and PC2) that we used for further analysis. 

Second, for each bird species that we took colour measurements, we scored the following variables: the light of the environment that the particular species inhabits (light environment), hunting strategy, body size, the system of parental care, and presence and absence of territoriality. The values for each variable were scored from the following resources: 1. EltonTraits database (Wilman et al., 2014) (Body size); 2. Publications on ecology and behavioural traits of Coraciiformes (Fry et al., 1992) (Light environment, hunting strategy, parental care, territoriality); 3. Online resources: Handbook of the Birds of the world alive (Light environment, parental care, territoriality). The information from these resources was processed by using already established categories for each of these variables known from multiple publications on the ecology and behaviour of birds. 

Usage notes

Microsoft Excel spreadsheet is necessary to open the data files, and R programming language is necessary for further processing. 


Royal Society, Award: Enhancement Grant RGF\EA\180155

European Research Council, Award: 615709 Project ‘ToLERates’

Royal Society, Award: University Research Fellowship UF120016 and URF\R\180006

Leverhulme Trust, Award: Early Career Fellowship ECF-2018-101

Natural Environment Research Council, Award: Independent Research Fellowship NE/T01105X/1