New York Times
POSTED: 1:30 a.m. HST, Feb 5, 2013
In 1855, Charles Darwin took up a new hobby. He started raising pigeons.
In the garden of his country estate, Darwin built a dovecote. He filled it with birds he bought in London from pigeon breeders. He favored the fanciest breeds — pouters, carriers, barbs, fantails, short-faced tumblers and many more.
"The diversity of the breeds is something astonishing," he wrote a few years later in "On the Origin of Species." — a work greatly informed by his experiments with the birds.
Pigeon breeding, Darwin argued, was an analogy for what happened in the wild. Nature played the part of the fancier, selecting which individuals would be able to reproduce. Natural selection might work more slowly than human breeders, but it had far more time to produce the diversity of life around us.
Yet to later generations of biologists, pigeons were of little more interest than they are to, say, New Yorkers. Attention shifted to other species, like fruit flies and E. coli.
Now Michael D. Shapiro, a biologist at the University of Utah, is returning pigeons to the spotlight.
In an article published online last week by the journal Science, an international team of scientists led by Shapiro reports that it has delved into a source of information Darwin didn't even know about: the pigeon genome. So far, they have sequenced the DNA of 40 breeds, seeking to pinpoint the mutations that produced their different forms.
The scientists are following Darwin's example by using the birds to find clues to the way evolution works in general. They are particularly interested in the mutations that produce radically new kinds of anatomy.
"Pigeons are an ideal way to look at these things," Shapiro said.
The new work supports Darwin's original claim that all pigeon breeds descend from the rock pigeon, whose range stretched from Europe to North Africa and east into Asia.
"It's a brilliant bit of investigative science, the type of research that hopefully will come to define the genomic era," said Beth Shapiro (no relation to Michael), an evolutionary molecular biologist at the University of California, Santa Cruz.
Archaeologists have speculated that rock pigeons flocked to the first farms in the Fertile Crescent in the Middle East, where they pecked at loose grain. Farmers then domesticated them for food.
Later, humans bred the birds to carry messages. By the eighth century B.C., Greeks were using pigeons to send the results of Olympic Games from town to town. Genghis Khan used pigeons to create a communication network across his empire in the 12th century A.D.
Eventually, people began breeding pigeons simply for pleasure. Akbar the Great, a 16th-century Mughal emperor, always traveled with his personal colony of 10,000 pigeons. He bred some of the birds for their ability to tumble through the air, and others for their extravagant beauty.
Michael Shapiro and his colleagues have been able to work out the genealogy of these breeds. They found, for example, that fantail pigeons, one of Akbar's favorite breeds, are closely related to breeds from Iran. Shapiro suspects that their kinship is a result of trade along the Silk Route between the Mughal Empire and Persia.
Some of these breeds would escape from their owners and mate with wild rock pigeons. As a result, Shapiro and his colleagues have struggled to find a pigeon with "pure" rock pigeon DNA. In search of wild birds, they sampled the DNA of pigeons from remote islands off the coast of northern Scotland. "If there's going to be any truly wild pigeons left, those are going to be a good place to look for them," he said.
The Scottish pigeons turned out to be closely related to Modena pigeons, an old Italian breed that may have interbred with the ancestors of wild pigeons in Scotland. Or perhaps Modena pigeons were domesticated directly from wild ancestors, rather than another breed. "We just can't distinguish between the two possibilities yet," Shapiro said.
European colonists brought their domesticated pigeons to the New World, where they raised them once more for food, messages and diversion. Thomas Jefferson designed a grand dovecote for Monticello, complete with pillars. Some of America's tame immigrant pigeons escaped yet again and evolved into a new population of feral pigeons — the ones that thrive in American cities.
"It looks like European and North American ferals are quite distinct," Shapiro said.
Like Darwin, Shapiro came late to the world of pigeon breeding. From 2001 to 2006, as a postdoctoral researcher at Stanford, he studied how stickleback fish in Canadian lakes evolved into strikingly different shapes in just a few thousands of years. While giving a talk for the position at the University of Utah, he had waxed poetic about how some of the fish had lost their armored spikes.
His host was not impressed.
"He said, ‘You think sticklebacks are diverse?"' Shapiro recalled. "And he plunked down an 800-page encyclopedia on pigeon breeding and said, ‘Take a look at this."'
Shapiro started leafing through the book.
"I knew a little bit about pigeons from Darwin, but this was insane," he recalled.
When he got the job in Utah, Shapiro decided to split his lab's efforts between fish and pigeons. He set out to discover what Darwin could not: the genetic basis of the birds' evolution.
When he explained the project to pigeon breeders he met on a visit to the Utah State Fair, they allowed him and his colleagues to draw blood from their birds to get their DNA. Before long, Shapiro was following Darwin's steps and raising pigeons of his own at the university, crossing the breeds to produce hybrids.
Once he and his colleagues worked out the genealogy of pigeons, they could then investigate how they had evolved into so many different forms. To begin this stage of the project, Shapiro picked out a particularly extravagant trait to study: head crests.
"There are many different kinds of crests," he said. "Some birds just have just a little peak, some have what looks like an inverted shell, some have a mane, and some have their entire head engulfed in feathers."
Shapiro and his colleagues have found that the closest relatives of crested breeds are uncrested breeds. In other words, pigeon breeders produced crests on the birds on five separate occasions.The scientists compared the genomes of the crested pigeons with one another, as well as with other pigeons and with chickens, turkeys and other species. They hunted for mutated genes unique to the crested breeds, and found that all of them shared precisely the same mutation in precisely the same gene, EphB2.
Bird embryos develop placodes, little disks of tissue on their skin from which feathers will grow. The scientists found that in ordinary pigeons without crests, EphB2 became active on the bottom edge of the placodes; in crested pigeons it was active on the top edge.
The experiment suggests that EphB2 tells the placode which way is up. In most pigeons, it instructs the feathers to grow down the neck; but the mutation changes the location where EphB2 switches on, effectively turning the feathers upside down and producing a crest.
"They grow the wrong way," Shapiro said. "They're even pointing the wrong way in the embryo, before they become feathers."
The new research suggests that the crested version of EphB2 arose in a surprising way. It mutated only once, rather than five separate times.
Shapiro came to this conclusion in part because he found that it takes two copies of the mutant gene to reverse the feathers. When the mutation arose, it was passed down invisibly from pigeon to pigeon. Only when two carriers happened to mate did they suddenly produce a crested chick.
Adam Boyko, a Cornell geneticist who studies dogs, has found similar results in his own research.
Several dog breeds have short legs, for example, but only a single mutant gene is responsible for the change. Like the crested-feather mutation, it worked its way into each of the short-legged breeds. "There's clearly a parallel," Boyko said.
Shapiro is moving ahead with studies on the many other traits of pigeon breeds to see if this pattern is an exception or the rule.
"The more examples that we have," he said, "the more we can understand what the general trends in evolutionary change are."