Adaptation, constraint, and chance in the evolution of flower color and pollinator color vision

Abstract

Anciently the teaching was that nothing would have been created that did not have a definite purpose, and more recently it has been that natural selection would eliminate anything that did not serve an equally definite purpose. … the assumed relation between the colors of flowers and the … pollinating insects is such a classic … Apparently there is something about the internal mechanism … that makes it difficult for a rose to be blue. … therefore, the use of the idea of natural selection to explain the absence of blue roses in nature is not only not necessary but it is not justified … It would be much better for the rose to be blue. F. E. Lutz (1924) We commonly think that biological signals and receivers are mutually tuned to one another. Flower colors and pollinator color vision are not exceptions. The diversity of flower colors and the differences in color vision between different classes of pollinators make speculations about their mutual adaptation tempting. Yet close inspection reveals that we know very little about evolutionary changes in flower color induced by selection pressures related to pollination, nor is there much evidence to show that color vision systems of pollinators have been tuned to flower color. We shall review cases where we think such changes have occurred, and other cases where they have not, even where a purely adaptationist scenario would predict evolutionary tuning.