Bird species in which one male can mate with many females tend to have more colorful males. But the promiscuity has an even stronger effect on females, making them drabber. That’s one of the more surprising conclusions in a new study of more than half of all living species of birds, which also reveals that a bird’s size and breeding location has a strong influence on the extravagance of its plumage.

“This paper is one of the most ambitious comparative studies ever conducted,“ says Geoffrey Hill, an ornithologist at Auburn University in Alabama, who was not involved with the work. But Richard Prum, an ornithologist at Yale University, says the paper is flawed because the team relied on pictures of birds in a book rather than observing them in the wild. “You couldn't study animal pheromones with scratch-and-sniff recreations."

Most scientists believe that bright colors signal good health or a great immune system. But why are some bird species more colorful than others? That’s been tough to resolve because it is hard to quantify how colorful a plumage is, says Bart Kempenaers, an ornithologist at the Max Planck Institute for Ornithology in Seewiesen, Germany. “How do you compare bright red with bright blue or yellow? That is the problem we had to solve."

Kempenaers and his colleagues tried a new approach: scanning pictures. The scientists focused on passerine birds, a group that makes up more than half of all known bird species and that is sometimes known as perching birds for their arrangement of toes—three pointing forward, one back. The researchers scanned illustrations in the Handbook of the Birds of the World, the only book covering every known living bird species, and then used a computer program to quantify how colorful each bird’s plumage is.

The tricky part was getting just one number that they could compare across species. For each bird, the scientists looked at six different patches of feathers (nape, crown, forehead, throat, upper breast, lower breast) and then identified the 1% of birds that were closest in color in the same patch. The more males, the higher the score for that patch. The researchers then calculated the average of the six patch scores for each bird. In essence, the scientists measured how "malelike" a bird appeared. But because male birds, in general, tend to be more colorful, that measure also works as a measure of how colorful a bird’s plumage is.

Analysis of the data yielded several trends: Larger birds are more likely to be colorful, possibly because they are less likely to be eaten by predators and can afford to be conspicuous. Tropical birds also tend to be more colorful, an observation already made by Charles Darwin. “We can’t say with any certainty what’s driving it. But we can say with certainty it is a very strong and real trend,“ says James Dale, an ecologist at Massey Unviersity. Albany, in New Zealand, and one of the authors of the paper published online today in Nature.

The authors noticed one more trend: In species where males mate with more than one female (called polygyny), male birds tend to be more colorful than in monogamous species. This was already known and it is seen as the result of strong competition between males for females. But the authors found that polygyny had an even stronger effect on females: It made them drabber. In monogamous species, males also get to choose females, so there is some sexual selection pressure on them to appear more beautiful. But males in many polygynous species basically take whatever they can get, Kempenaers says. “In these species, sexual selection is acting only on the men. For females there is just natural selection and that favors an inconspicuous plumage."

Tim Caro, an expert on coloration in animals at the University of California, Davis, says the paper is interesting because it looks at female ornamentation as well as male. “Usually most attention is focused on ornamented males," he wrote in an email.

Yet Prum, who has studied the evolution of avian plumage coloration, says that the generalizations arrived at in the study are meaningless, because evolution acts on individual lineages in different ways. For instance, he points out that the biggest passerine bird, the lyrebird from Australia, is large and polygynous, yet both males and females are also drab.

Prum also argues that studying avian plumage color off prints is “scandalous," because there are some aspects of bird coloration that birds can see, but humans cannot. These can only be captured by using a method called spectrometry on live birds or museum specimens, he says.

The authors did measure more than 500 bird specimens in museums to test their technique. “The results were very similar whether we used handbook plates or museum skin data," Dale says. Some information on color will not be picked up by the technique, he acknowledges. “But we compensate for that by having a very large sample size so that the results we get are still very biologically relevant."