Same colors for different functions: implications for the evolution of carotenoid-based ornamentation

<p><span>Sexual ornamentation is often assumed to be costly, allowing honest signaling of individual quality, and carotenoid-based colors were proposed to bear significant costs. If carotenoid-based colors are costly to produce, sexually-selected signals should use more concentrated carotenoid pigments and have more saturated color than non-sexual signals, where honesty-guaranteeing costs are not required. We tested this prediction comparing carotenoid-based colors across canaries, goldfinches and allies, because many of these species </span><span>use </span><span>yellow plumage as sexual ornamentation, but also have yellow rumps that appear to be non-sexual flash marks</span><span>. Only in the breast, but not the rump, was there an asymmetric </span><span>co-distribution of male and female color saturation, with males </span><span>similarly or more saturated than females, indicating evolution of breast color by sexual selection. Yellow was not consistently more saturated in the breast than in the rump, and the co-distribution of rump and breast color saturation indicated that saturated rumps can persist irrespective of breast color. This challenges </span><span>the assumption that carotenoid-based colors bear significant costs. The use of carotenoid coloration as sexual signals in this clade may instead be due to social costs, cost-free index mechanisms for signaling quality, and/or socially-monogamous species evolving low-cost signals to mostly discriminate against low-quality mates.</span></p>