Introduction
Coloration is arguably one of the most fascinating traits in evolutionary biology (Salis, Lorin, Laudet, & Frédérich, 2019). Maintenance and evolution of colour polymorphism is linked to pigment availability, pigment distribution and cell types that either produce, contain or transport pigments across cell layers, such as keratinocytes, melanophores, xanthophores and iridophores (Hofreiter & Schöneberg, 2010). The molecular basis of coloration has been extensively studied across taxa, revealing a multitude of loci involved in the production of colour derived from a) the interplay between pigmentation availability-distribution and b) existence of taxa-specific molecular pathways for colour production (Hofreiter & Schöneberg, 2010). For instances, while mammals owe their colour of hair or skin to a deposition balance between eumelanin and pheomelanin, which are forms of melanin, other taxa might manifest coloration via melanin deposition, carotenoid deposition and structural arrangement of cell layers, i.e. structural colour (Hofreiter & Schöneberg, 2010).
Melanin-based pigmentation is the most ubiquitous colour-producing pigment in extant birds (Roulin, 2014). The melanin production pathway is complex and evolutionarily conserved but centered on the POMCgene (which encodes for the protein proopiomelanocortin) (Ducrest, Keller, & Roulin, 2008). This protein is cleaved into several peptides – melanocortins – that bind to up to 5 melanocortin receptors (MC1R-5R ) each involved in specific physiological functions, including melanin-production (MC1R ) (D’Alba & Shawkey, 2019; Ducrest et al., 2008). Though the pleiotropic nature of the POMCgene has been posited to explain associations between melanin-based pigmentation and other phenotypic traits, i.e., melanin-associated phenotypes, molecular evidence is largely absent (McNamara et al., 2021; Luis M San-Jose & Roulin, 2018). What has been well described is structural genetic variation at MC1R and respective regulatory regions affecting melanin-based pigmentation in several bird species including domestic breeds (Kerje, Lind, Schütz, Jensen, & Andersson, 2003; Liu et al., 2023; Mundy, 2005).However, lack of associations has also been reported particularly after correcting for the effect of population structure (Avilés et al., 2019; Hoffman, Krause, Lehmann, & Krüger, 2014; MacDougall-Shackleton, Blanchard, & Gibbs, 2003).
Arguably, while most colour polymorphisms identified in birds tend to be protein-coding changes, the possibility for a wide variety of more quantitative loci defining colour do exist (Johnsson, Jonsson, Andersson, Jensen, & Wright, 2016; Roulin & Ducrest, 2013). The array of quantitative colour-loci is evident among bird taxa, because in those organisms coloration is linked to plumage. Feathers themselves have a variety of different functions that first emerged duringSauropsida evolution, such as flight, protection, and mechanical insulation via keratin-based micro-structures:(barbules, micro-barbules and hooklets) (Benton, Dhouailly, Jiang, & McNamara, 2019). Melanin deposition further enhances a feather’s properties: melanin-packed melanosomes (specialized organelles in animal’s cells where melanin synthesis and storage occur) grants thickness and resilience to torsion/tension, provides chemical defences by oxidizing or reducing free oxygen radicals and participate in thermoregulation by absorption of light and conversion into heat (D’Alba & Shawkey, 2019; Field et al., 2013). Given the multiple functionalities of the feather and melanin, it is plausible to expect the emergence of evolutionary trade-offs when environmental conditions favour divergent phenotypes (Shoval et al., 2012; Terrill & Shultz, 2023). Not surprisingly, trade-offs related to feather/melanin functions have been shown to occur in birds in signalling and social recognition (Sheehan & Bergman, 2016), thermoregulation (Galván, Rodríguez‐Martínez, & Carrascal, 2018), and resource allocation particularly in the production of the costly pheomelanin pigment (McNamara et al., 2021). From a molecular point of view, putative trade-offs would translate in co-variation between melanin-based phenotypes and loci involved in functions that mitigate the colour effect.
About one-third of owl species are colour polymorphic and typically harbor a higher ratio of the light colour morph in higher latitudes consistent with Gloger’s rule (Galeotti, Rubolini, Dunn, & Fasola, 2003; Passarotto, Rodríguez‐Caballero, Cruz‐Miralles, & Avilés, 2022). Particularly, the tawny owl is a nocturnal bird of prey exhibiting a melanin-based colour polymorphism ranging between pale-grey and a pheomelanin-dominated reddish-brown (Brommer, Ahola, & Karstinen, 2005). The inheritance pattern of colour polymorphism in this species is apparently consistent with a Mendelian one locus-two allele system, is independent of sex and age and is mediated by the irreversible accumulation of melanin on plumage feathers during feather development (Brommer et al., 2005). On a large geographical scale, the grey morph is more prevalent than the brown in environments with colder and less rainy winters and in the northern boreal zone. The high prevalence of the grey tawny owl morph in northern latitudes may reflect a selective advantage via background matching (crypsis) under snowy conditions to increase protection against intra-guild predation and anti-predator mobbing (Koskenpato, Lehikoinen, Lindstedt, & Karell, 2020). Indeed, morph frequencies over ecological timescales were shown to be skewed towards grey as a function of severe, snow-heavy winter conditions in higher latitudes (Brommer et al., 2005; Karell, Ahola, Karstinen, Valkama, & Brommer, 2011). Snow-covered environments might pose two potential costs for the maintenance of colour polymorphism in tawny owls. The first derives from the fact that reduced melanin deposition confers a putative selective advantage in the form of crypsis, decreased pigment-based thermoregulation is plausible since melanin-rich morphs are expected to absorb more sunlight and thus reduce the cost of homeothermy – though daylight perching might also occur in the shade (Galván et al., 2018; Roulin, 2014). The second derives from the scarcity of resources during extreme winters which might hinder the production of pheomelanin in the spring breeding season due to the unavailability of protein (Britton & Davidowitz, 2023). In relation to the former, Koskenpato et al (2016) found a higher density of insulative barbules in feathers of grey (no pheomelanin) tawny owls (Koskenpato, Ahola, Karstinen, & Karell, 2016), which to the best of our knowledge remains the only evidence suggesting a trade-off between form (colour) and function of feathers.
In this study, we present the first pair of draft assemblies of the tawny owl genome coupled with a holistic genome screening approach to identify genome-wide structural variation associated with colour phenotypes. We capitalized on a decade of data collection from a tawny owl population comprised of 96 progenitors and their respective offspring (for a total of 370 individuals,) all of which have been individually phenotyped for colour. Pedigree availability allowed us to undertake a series of genome-wide association studies while controlling for heritability and population structure. Our objectives were (1) to identify the genetic basis of colour polymorphism in tawny owls and (2) to inspect the molecular basis of putative melanin-phenotype associations by associating the genomic information with the extensive phenotypic ecological data available for this study system in southern Finland. Overall, identifying the genomic basis of colour and colour-associated phenotypic traits is fundamental to understand evolution and maintenance of colour polymorphism.