In the animal kingdom, there is incredible variation in visual perception. What an animal sees depends on the structure of its retina and its neural visual processing system. Most insects can see ultraviolet, blue and green light, but there is wide variety among arthropods; mantis shrimp eyes have up to 12 different channels of color, revealing the ultraviolet spectrum and polarized light. The ancestor of living vertebrates could likely detect red, purple, blue and green — an ability that was maintained in lizards, birds, lampreys and lungfish, among other vertebrate groups, Wiens said. But some components of color vision have been lost over evolutionary time. Hagfish can’t detect red. Sharks can’t see blue. Human eyes have three photoreceptor cones that allow us to make out blues, greens and reds, but dogs and rabbits have only two cones, which reduces the number of shades they can distinguish.
Wiens and Emberts’ data supports the hypothesis that color evolved for some as-yet-unknown reason before any of these flashy signals. “It was color vision first, then fruit, then flowers, then warning signals and then sexual signals,” Wiens said.
Coming and Going
The researchers’ effort to reconstruct deep time is admittedly imperfect. Colors don’t readily fossilize, and when they do, scientists can’t infer the color’s function unless the animal has living descendants. And for all the data they involve, evolutionary trees are inherently speculative. Some traits can evolve multiple times in different lineages. For example, juniper berries and blueberries are both blue, but their ancestors may have developed that coloration separately. Other traits can come and go, like the lizards’ blue belly patches. If we know that signals can disappear and reappear over millions of years, it’s hard to be certain that a common ancestor actually possessed that shared trait.
“This evolutionary lability has the effect of blurring whether a color adaptation existed in deep time or not,” Allen said. “If, for example, a lineage gains or loses warning color once every million years, it is very difficult to infer from the traits of current species whether an ancestor living hundreds of millions of years ago had warning color.”
That’s why using phylogenies to date the origin of a function has inherent uncertainties. “Looking at what’s around nowadays doesn’t tell you very much because it’s just coming and going,” said the neuroscientist Daniel Colaco Osorio, who studies animal vision at the University of Sussex and was not involved in the study. But Wiens insists that the only way we can know that traits disappear and reappear is by using evolutionary trees to test this kind of hypothesis. This is the best method we have to peer into the evolutionary past from the present, he said. “How does one claim [that traits come and go] without doing a study like ours?”
To be clear, there was color in the world before color vision. Plant leaves, for example, reflect green light even if there are no eyes to see it. In 1999, Osorio studied color vision in chicks and suggested that it serves a more general purpose. “It could just be recognizing objects or navigating around the place,” he speculated. A prevailing theory was popularized in 2000, when Vadim Maximov proposed that color vision evolved to aid vertebrates in low-light aquatic conditions. The presence of two classes of photoreceptors, he argued, helped reduce the “flicker” beneath the surface of shallow water, which helped aquatic creatures chase prey and avoid predators. That would explain why the building blocks of color vision arose just after the active predatory lifestyle was mastered, but well before there was a more obvious use for it.
Color signals maintained by living things across evolutionary time are of a different character. In 2019, Osorio suggested that a vivid color, whether it’s a pure pigment or a reflective structure, takes action to organize. They are therefore evidence of work against the forces of entropy; it’s generally something an organism evolved for a reason. “If you empty the contents of your vacuum cleaner bag, it’s kind of gray because everything’s mixed together,” he said. “If you have a structure with meaning or purpose, that can be indicated by having a bright or pure color … which isn’t particularly tied to the meaning of the signal.” His research has demonstrated mathematically that what looks like a vivid, pure color to one animal is likely to look vivid and pure to another, regardless of their visual system.
Some questions are ultimately unknowable — but can still be useful or enjoyable to ask. Wiens and Emberts’ review is a step toward understanding why the natural world is so colorful. Plus, Wiens found that over the past 100 million years, there’s been an explosion in warning and sexual signals driven by signaling between birds, lizards and fish. He believes that this trend might continue — which suggests nature is on track to get even more dazzling.