André A. Fernandez and Molly R. Morris have revisited a question that evolutionary biologists just can't let go of. Many species of primate have trichromatic colour vision. But what factors drove its evolution?

The authors have tested the 'pre-existing-bias hypothesis' (Am. Nat. doi:10.1086/518566; 2007). This holds that the evolution of a trait such as colour vision might have evolved to meet one particular need, but was subsequently favoured by selection for another.

A long-standing idea is that trichromatic vision is of great benefit in foraging — in allowing red or orange food items to be seen more easily in green foliage. Another is that its genesis lies in promoting sexual selection — that intense colour, in particular red as skin or fur coloration, evolved for example as a signal of sexual state or prowess. Fernandez and Morris set out to provide a statistical test of the possibility that selection of this latter putative function was due to a pre-existing bias.

Their approach was to mine the literature to produce 'ancestral state reconstructions' of the incidence among primate groups of colour vision, fur colour, skin colour and form of mating system. The thinking behind the inclusion of this last characteristic was that selection for colourful signals is likely to be greater in gregarious species. These reconstructions were then subjected to tests of whether red fur and skin were more likely to have evolved in the presence of colour vision and gregarious mating systems.

The conclusion that emerges is that trichromatism preceded a tendency towards red coloration and gregarious mating systems — that is, that there was indeed a pre-existing bias towards red (probably, think the authors, for more efficient foraging) that then prompted the coevolution of red skin and fur through sexual selection.

There are obviously plenty of exceptions to the correlation between trichromatism and red coloration. Explanations for such exceptions include lack of suitable genetic variation in the species concerned, lack of the appropriate diet or predator pressure.

Fernandez and Morris add that their research may tie in with work on the evolutionary history of the vomeronasal organ. This is a sensor for scent signalling, and there is evidence that its functional loss in Old World monkeys coincided with the advent of colour vision — prompting the idea that chemical signalling gave way to visual cues.