Here's a link to an interesting and frequently-cited article concerning human dichromacy. There is a remarkably high incidence of dichromacy in the human population (up to 2%). The authors examine a potential explanation for this trend and show that dichromats can detect color-camouflaged texture patterns more quickly than trichromats. They explain that trichromats turn first to color differences when 'segmenting' visual inputs, a strategy that can impede their recognition of texture patterns that lie beneath the color differences. This article introduces interesting ideas about whether benefits such as camouflage-detection help to maintain visual polymorphism in the human population, or whether non-adaptive genetic factors such as unequal crossing-over at the red-green locus are responsible for the high incidence of dichromacy in humans.