Bee color vision is optimal for coding flower color, but flower colors are not optimal for being coded - Why?

Model calculations are used to determine an optimal color coding system for identifying flower colors, and to see whether flower colors are well suited for being encoded. It is shown that the trichromatic color vision of bees comprises UV, blue, and green receptors whose wavelength positions are optimal for identifying flower colors. But did flower colors actually drive the evolution of bee color vision? A phylogenetic analysis reveals that UV, blue, and green receptors were probably present in the ancestors of crustaceans and insects 570 million years ago, and thus predate the evolution of flower color by at least 400 million years. Ln what ways did flower colors adapt to insect color vision? The variability of flower color is subject to constraint. Flowers are clustered in the bee color space (probably because of biochemical constraints), and different plant families differ strongly in their variation of color (which points to phylogenetic constraint). However, flower colors occupy areas of color space that are significantly different from those occupied by common background materials, such as green foliage. Finally, models are developed to test whether the colors of flowers of sympatric and simultaneously blooming species diverge or converge to a higher degree than expected by chance. Such effects are indeed found in some habitats.