Although city dwellers may not want to admit it, humans and pigeons (also known among detractors as "flying rats") have been intertwined since ancient times. We domesticated pigeons between 3,000 and 5,000 years ago, maybe even before we let cats inside our homes. Nowadays there are more than 350 distinct pigeon breeds, each created by artificial selection; many of which so tailored to human preferences they could never survive in the wild. They are so varied, Charles Darwin once remarked, that if they had been discovered in nature, they would have been considered to be several separate species.

Slide Show: Genetic Classification Reveals Pigeons' Exceptional Diversity

Yet despite the birds' prevalence throughout human societies and across history, neither breeders nor biologists were sure how the breeds were related to one another; classification schemes based on morphology and historical accounts varied among reference books. To clarify the family tree of pigeon breeds, Mike Shapiro, a biologist who studies the genetics of diversity at the University of Utah, undertook the first ever genetic classification of domestic pigeons, published in the January 19 issue of Current Biology.

Trevor Price, who studies bird speciation at the University of Chicago, says the paper "provides an essential framework on which to investigate the origin of diversity among pigeons. In that sense, it is an incredibly important study." Not only are pigeons fascinating in their own right, they provide a general model of evolution by natural selection.

With the help of hundreds of pigeon breeders, Shapiro's team collected blood and feather samples from more than 350 birds representing 70 different breeds. "We went to shows and asked to take blood samples from their prized birds—and to our delight and surprise, most breeders let us," Shapiro says. "Pigeon hobbyists are interested in their birds from every perspective, and most of them are also intuitive geneticists. They taught us a few things about pigeon genetics."

In fact, it turned out that pigeon fanciers had done a pretty good job of sorting the birds into breeds; when Shapiro's group used DNA markers to determine which breeds were mostly closely related, the DNA revealed only a few surprises (discussed in the slide show). Perhaps more interestingly, the team found that breeds that were genetically similar weren't always morphologically similar, and vice versa. This mismatch implies that some traits—such as feathered feet or head crests—may have evolved independently in different breeds, based on pigeon fanciers' preferences.

"It's fairly clear that these traits arose more than once in different breeds, and it would be interesting to know whether that was because of independent mutation or interbreeding," Price says. "We wouldn't even be able to ask that question without Shapiro's analysis."

The relationships between breeds are clearer now, but some genetic distinctions are difficult to make simply because pigeons diversified so rapidly and can still interbreed. In some ways, the paper raises more questions than it answers. Did domestic pigeons' genetic diversity originate in the wild or did humans shape it? Is their flexible morphology based on changes in the actual DNA structure, or changes in its expression? In breeds where traits appear to have arisen independently, Shapiro says he wants to see if the same genetic mutations are causing the trait.

The abundant variation found within domestic pigeon breeds may also turn out to be a tool for evolutionary geneticists. "In an ideal world, if we wanted to determine which genes are responsible for beak size and shape, we could set up crosses between a toucan and an owl," Shapiro says. In pigeons, that variation is recapitulated between the miniscule beak of the African Owl pigeon and the downward-curving schnoz of the Scandaroon pigeon. "Owls and toucans can't interbreed, but pigeons can."