Age, sex and beauty: methodological dependence of age- and sex-dichromatism in the great tit Parus major

Carotenoid-based colour expression is frequently involved in sexual dichromatism, particularly in bird plumage, suggesting a role in sexual selection. Despite much work on expression of the carotenoid-based ventral plumage coloration of the great tit (Parus major), which represents a popular model in evolution and ecology, a consensus on even the most basic demographic patterns of variation (e.g. age and sex differences) is lacking. This may reflect the use of variable methods for analysing colour variation, although what is not clear, either in this case or in general, is the extent to which these alternative methods are equally effective at describing age- and sex-related dichromatism. Using data obtained over 4 years from a large sample of free-ranging great tits, we examined how colour-scoring methodology influences estimates of age- and sex-related dichromatism. We compare: (1) principal components analysis-derived scores; (2) tristimulus colour variables; (3) a visual model-independent, carotenoid-focussed colour score; and (4) two colour scoring methods based on avian visual models, examining how they assess colour variation with respect to age and sex to determine how methodology may influence results. We demonstrate clear age- and sex-dependent expression of this colour trait, both in our own data and in meta-analyses of results from great tit populations across Europe, and discuss the merits of the various colour scores, which yield very different estimates of the extent of age- and sex-dependent dichromatism. We show variation is likely to be visible to conspecifics and propose a novel, visual model-derived scoring system for describing variation in carotenoid-based colour patches, where the perceived signal is divided into independent chromatic and achromatic components, in line with current understanding of visual perception. The present study highlights the impact of colour-scoring methodology and shows that, as novel measures continue to be developed, researchers should consider carefully how they quantify colour expression. (C) 2010 The Linnean Society of London, Biological Journal of the Linnean Society, 2010, 101, 777-796.