If you really want to know about birds, you have to consider the crocodile.
That point was driven home this week with the release of the genomes of 45 bird species, which reassigned some perches on the avian evolutionary tree and included some seemingly odd bedfellows.
Down near the roots of that avian tree lies a mysterious ancestor that was decidedly more terrestrial and terrifying than the finch or the wren.
The archosaur, or so-called "ruling reptile," roamed Earth about 250 million years ago, and "was something that was very reptilian, very early-dinosaur-ish, and then it evolved into modern-day crocodiles and birds," said David Haussler, Scientific Director of the UC Santa Cruz Genomics Institute, a coauthor of several studies that came out of the avian genomics effort.
"So it really is the proper dinosaur ancestor," Haussler said. "And birds and crocodiles are the proper descendants of this ancestor."
Haussler isn't a fossil-digging researcher. He digs through genetic code. So does John McCormack, an Occidental College biologist who usually is plenty busy curating a collection of some 65,000 Mexican birds at Moore Lab of Zoology on the college's Los Angeles campus.
But both researchers are keenly interested in a kind of living molecular fossil -- small strands of DNA, the code of life, that are shared among a wide array of species.
"These markers are very nice for doing comparative genomics, because they're so conserved. They're easy to find among organisms that are very distantly related," said McCormack. "We can find them across all of these genomes, and use them to build a phylogeny -- an evolutionary history."
That's where the modern saltwater crocodile, American alligator and Indian gharial come in. Those modern crocodilians are still crawling around with much of the DNA they inherited well before dinosaurs ruled and evolved into birds. That's why McCormack and Haussler helped map out the modern crocodile genome, along with those of living birds. Their work was among 28 research papers published online Thursday, based on a four-year genome mapping effort.
They found the crocodile had the slowest rate of molecular change of any known vertebrate genome.
"The DNA in the modern-day crocodile has changed a lot less, versus the archosaur, than it has in birds," said Haussler.
By comparison, bird DNA took flight. Their pace of molecular evolution accelerated, and within about 15 million years of evolution, birds radiated out into the bulk of the existing 36 modern avian orders.
"There's a deep split back in modern birds, and it leads to two lineages," McCormack said. "Each of those lineages goes on to split into separate water bird and land bird lineages, in parallel. That's totally new."
The resorting of species resulted in large part from an analysis of the "ultraconserved" bits of DNA, much of it between true genes. These segments don't code proteins that ultimately lead to what makes the difference between a flamingo and a hummingbird. Most of them appear to regulate genes.
Evolutionary trees based on this type of DNA uprooted those that were drawn based only on a relative few coding genes. That left a consensus tree that is different than previous versions.
"It would be great if every single species consisted of a lot of individuals all with the identical genome, and then a mutation happens and you have two species and they get slightly different genomes over time," Haussler said.
But life doesn't work out that way, in part because populations already will have a lot of genetic variety. And that variation in genes can make it hard to sort out lineages.
Take hawks and falcons, McCormack said. In the old days, they were categorized as birds of a feather.
"Most markers in the genome will tell you that hawks and falcons -- even though they look somewhat similar and have similar lifestyles -- are actually not very closely related to one another," McCormack said. "However, if you look only at protein-coding genes, they seem closely related to one another."
In another example, hummingbirds look a lot like songbirds, based on protein-coding genes, McCormack said. The new tree places them closer to swifts or cuckoos.
"In the good old days, species were classified based on what you could measure," or traits, Haussler said. Now, "the actual place in the genome is like the ultimate molecular trait," he said.
The bird genome turned out to be slim as vertebrates go -- about 1-1.26 billion base pairs, which are the molecules that join together to make the lattice of DNA's double helix. Birds seem to have ditched a lot of DNA as they took flight, jettisoning the code for such things as teeth and a second ovary, the studies show.
Nearly all of the papers published Thursday dealt with these differences, in minute detail. But Haussler said the fundamental truth of genetic history is that a vast amount of DNA is shared among species -- from the basic functions of cells to larger body plans and the function of body parts.
"What's important are the regulatory regions that control when and where the genes come on during the development of the body and the maintenance of the adult body," Haussler said. "And they are more rapidly evolving than the actual standard genes."
Some fish, for example, have the genes for legs, research shows. They just don't have the code that makes fins turn to fingers.
When Haussler and others compared the human and mouse genomes, they found so much similarity they suspected their samples were cross-contaminated, he said.
"Regions of hundreds of bases of DNA were identical, without any change, over a period of close to 100 million years," Haussler said.
"You don't create an entirely new gene when you are adopting a new trait," he said. "Normally, you take the genes that are there and regulate them differently. This is how evolution primarily works. And that story gets played out again and again and again."
As the saying goes: Evolution is a tinkerer, not an inventor.
So, on some very fundamental molecular level, every bird is the Egyptian plover -- standing on the back of the crocodile.