Elevated testosterone levels during moult have contrasting effects on structural and carotenoid‐based plumage colours in Eurasian Blue Tits

Feather colours can be produced by the organization of feather microstructure, by pigmentation (mainly due to melanins and carotenoids) or both. The expression of feather colours is usually influenced by testosterone (T) levels through several interacting mechanisms. First, high T levels can negatively affect bird condition and hence the expression of condition‐dependent feather colours (e.g. carotenoid‐based colours). Secondly, high levels of T can slow moult progression, which in turn may result in brighter colours, as moult speed has been shown to be negatively correlated with the expression of feather colours. Thirdly, T can affect the bioavailability of pigments that are involved in feather colours. The effect of experimental manipulation of T levels may therefore influence, either positively or negatively, feather colour expression according to the relative importance of these mechanisms and the type of coloration involved. We experimentally investigated whether plumage coloration is affected by T in yearling Eurasian Blue Tits Cyanistes caeruleus. We implanted 11 males and 11 females with T (T‐birds) at the onset of their post‐juvenile moult and compared the spectral reflectance of their structural ultraviolet (UV)/blue colour (crown feathers and upper lesser wing‐coverts) and carotenoid‐based yellow colour (breast feathers) with those of control birds (C‐birds) that were sham‐implanted (12 males, 16 females). At the implant date, all the right lesser wing‐coverts were plucked in both experimental groups to test the effect of T on feathers grown exclusively during our treatment. After 40 days, the implants were removed. Two weeks before removing the implants, the mean T plasma levels were assayed (average 3.21 ng/mL in T‐birds and 0.33 ng/mL in C‐birds). In T‐birds, body moult progression was suspended, and moult was resumed 33 days after implant removal. Moult duration from implant date, excluding suspension, did not differ between treatments and controls. T had a negative effect on UV coloration only in plumage that started growing exclusively during the treatment (right lesser wing‐coverts), and not in other UV/blue plumage. In contrast, the carotenoid‐based yellow coloration of the breast increased in T‐birds, suggesting an immediate effect of T on the mobilization of carotenoids available for ornamentation. Despite sexual dichromatism, T treatment had similar effects in the two sexes. Our results highlight the importance of considering the effect of T on moult speed and possibly moult suspension in the interpretation of the results of studies based on T manipulation.