Carotenoid-dependent plumage coloration is associated with reduced male care in passerine birds

The immense diversity of plumage coloration exhibited by birds is the result of either pigments deposited in the feathers or microstructural arrangements of feather barbules. Some of the most common pigments are carotenoids, which produce bright yellow, orange, and red colors. Carotenoids differ from other pigments since birds cannot synthesize them de novo and must obtain them from the diet. Carotenoid pigments are usually associated with signaling and sexual selection, although they also have antioxidant properties and play a role in the immune response. Here, we hypothesize that carotenoid-dependent plumage coloration functions as a signal of a male's tendency to invest in offspring care because they play an important role in self-maintenance and may provide key information about individual quality, allowing females to obtain information about a male's tendency to invest in offspring care. Using phylogenetic comparative analyses across 349 passerine birds, we show that species that consume carotenoid-rich foods have more carotenoid-dependent plumage coloration than species with carotenoid-poor diets. In addition, carotenoid-dependent plumage coloration is associated with decreased male investment in offspring care. Our results suggest that investment in carotenoid-dependent plumage coloration trades off against male investment in offspring care and will likely have broad implications for our understanding of the ecological contexts that facilitate various evolutionary processes, such as sexual selection and signaling associated with plumage colors.

Methods
Data were collected from published sources, including both primary reference works (e.g., Olson and Owens 2005; Thomas et al. 2014; Liker et al. 2015) and compendia such as the Handbook of the Birds of the World Alive (del Hoyo et al. 2016; accessed between May and December 2016). 

Liker et al. (2015) compiled detailed information on relative investment by each sex in six components of avian parental care (nest building, incubation, nest guarding, chick brooding, chick feeding, and chick guarding (guarding and defending the brood post-hatching)) from reference books and published literature (references in Liker et al. 2015 updated in Gonzalez-Voyer et al. 2022). We focus here on only three components: nest building, incubation, and chick feeding, which are the ones for which information was available for the greatest number of species.

We used an existing dataset (Thomas et al. 2014) on the presence of carotenoid-dependent plumage coloration in all neornithine species. Presence (1) or absence (0) of carotenoid-dependent coloration was scored for all species based on plumage colors observed in illustrations taken from the Handbook of the Birds of the World Alive online (del Hoyo et al. 2016), using as a reference a set of species for which the presence of carotenoids in plumage had been chemically confirmed. We selected all passerine species from the list and ensured that it was the male who presented carotenoid-dependent coloration. As we were interested not only in the presence of carotenoid-dependent coloration but also in the amount expressed on the body of the species, we developed our scoring system following Owens and Hartley (1998) to obtain a quantitative estimate for the amount of area of the body with plumage with carotenoid-dependent coloration. For species for which Thomas et al. (2014) had confirmed the presence of carotenoid coloration, we estimated the amount of carotenoid-dependent coloration, whereas for those species without carotenoid-dependent coloration according to Thomas et al. (2014), we assigned a value of 0. We estimated the amount of carotenoids distributed throughout the body of the bird because the more pigments you need, the more expensive the signal is likely to be, and we assume that the selection pressure will be stronger. We only considered the colorations that typically depend on these pigments: bright reds, oranges, and yellows. To minimize bias, we selected three volunteers with birdwatching experience, but uninformed about the hypotheses we wished to test, to score the extent of each patch covered by carotenoid-dependent plumage coloration. Each observer was shown illustrations from the Handbook of the Birds of the World of the 177 passerine species that Thomas et al. (2014) classified as presenting carotenoid-dependent plumage coloration and asked to score patches with: if the carotenoid-dependent coloration covered the whole patch or a part of it; 0 if it did not show carotenoid-dependent coloration. The observers scored eight body regions per species: head, nape, back, rump, throat, chest, belly, tail, and wings. In the rare instances where the patch observers were meant to score was not visible in the images, we searched descriptions and pictures of the species to verify if they had carotenoid-dependent coloration in the patch. We scored these regions because they cover the whole body of the bird. In cases where multiple subspecies were illustrated, we scored coloration in the nominate subspecies. The continuous measure for carotenoid-dependent coloration was the sum of the scores for all patches for each species.

To assess the carotenoid content in the diet of each species, we developed a scoring system following Olson and Owens (2005), who scored carotenoid intake based on a coarse-scale index that ranks seven diet categories based on their carotenoid content: 1 for seeds, nuts, wood; 2 for nectar, pollen, sap, exudates, lerps; 3 for vertebrates; 4 for invertebrates; 5 for foliage, flowers, fungi; 6 for fruit; and 7 for algae, diatoms. We collected diet data from verbal descriptions in the Handbook of the Birds of the World Alive (del Hoyo et al. 2016). Since most species do not feed solely on a single group, we assigned relative values based on the importance of each group in their diet as described in HBW Alive. 

We used body mass as an estimate of the body size of the males of each species. We included body size in our analyses because it is strongly correlated with life history traits and could reflect possible restrictions for the development of plumage coloration in birds, such as metabolic rate and the ingestion of food with carotenoids. Body mass data were obtained from Liker et al. (2015) for almost all species. For species lacking data, we extracted it from Dale et al. (2015) and the HBW Alive. Log-transformed values of body mass were used in the statistical analysis.

We obtained the species’ breeding distribution from del Hoyo et al. (2016) for latitude.