Ultraviolet (UV) vision, first discovered in ants more than a century ago, is a major area of interest for behavioural ecology. Because, unlike humans, many animal species can see UV light, spectrometry in the UV has revealed fascinating signalling systems that remain hidden to the human eye. In his letter to Trends in Ecology & Evolution, Sinkkonen  argues that UV should also be considered in investigations on autumn colours, particularly when testing hypotheses that involve insect–tree interactions. Because UV vision is a generic feature of insect vision and has been confirmed for aphids , the main taxon of interest in the autumn colour debate, we agree with Sinkkonen that the role of UV needs to be explored in this area. Specifically, we need to know: (i) how much UV is reflected by autumn leaves and how much trees vary in UV reflectance; and (ii) how insects respond to UV leaf reflectance in autumn.
With regard to the first question, it is important to consider spectral reflectance in actual leaves rather than the spectral properties of isolated phytochemicals. The presence of UV-reflecting compounds in a leaf does not mean that the leaf itself will be UV reflecting, given the number of other leaf compounds present that might absorb UV light. In senescent leaves of Gingko biloba and Quercus robur, which Sinkonnen mentions as containing UV-reflecting compounds , maximal UV reflectances were found of only 10% and 9%, respectively (unpublished from Ref. ). Moreover, the UV-reflective compound in senescent G. biloba does not occur widely, especially in woody plants, and it would be misleading to extrapolate to trees as a whole. Indeed, an analysis of 2409 autumnal leaf spectra reveals that 99% have a maximal UV reflectance of 30% . By contrast, the low overall UV reflectance in autumn leaves means that the necessarily small differences among leaves would be relatively hard for insects to detect. In fact, a colour choice model developed from trapping migrant aphids in autumn found no effect of UV reflectance, when trap colours, mimicking the situation in leaves, reflected little UV . Further studies are needed to test more thoroughly how aphids respond to UV, but for low UV reflectance (ca.