New study shows animal family tree looking bushy in places
Dec. 22, 2005
Two decades ago, with the advent of methods to look at the family relationships of different organisms by analyzing DNA, scientists envisioned it would only be a matter of time before the various family trees for plants, animals, fungi and their kin would be resolved with genetic precision.
And while molecular methods have had enormous success in ordering some branches in the tree of life - mammals, for example - and have played a critical role in refining and correcting trees constructed on the more traditional means of the appearance of organisms, the tree of animals remains fuzzy.
Now, scientists may know why this is so. Writing this week (Dec. 23, 2005) in the journal Science, a team of UW-Madison scientists led by Antonis Rokas, now of the Broad Institute of MIT and Harvard, suggests that a branch-by-branch account of animal relationships over a vast expanse of time is difficult to reconstruct because early animal evolution occurred in bunches.
"In general, we'd like to know who's related to whom, and the pattern of the branches of the tree of life," says Sean Carroll, a Howard Hughes Medical Institute investigator at UW-Madison and the senior author of the Science paper.
But 500 million years of animal history on Earth is a lot of ground to cover, Carroll laments, and now it seems that the periodic, frenetic bursts of evolution that occurred at certain times in the distant past make sorting out animal relationships - the branches on the tree - extraordinarily difficult.
"It turns out that early in the origin of many types of animals, there were a lot of branching events in a short period of time," Carroll explains. Those type of episodes at key junctures in life's history - for example, the rise of complex animals or the migration of vertebrates from the sea to land - make the animal tree look very bushy and very murky, Carroll's group reports.
In the new Wisconsin study, which was also co-authored by Dirk Krüger of UW-Madison, massive amounts of molecular data for many animals were used to try to generate a clear picture of the animal tree.
"But instead of a tree, we got a bush where many branches sprout close together," Carroll says.
The group used the same approach to resolve the family tree of fungi, organisms that originated about the same time as animals. "In contrast to the animals, the tree of fungi, resolved neatly," Rokas explains. "The difficulty we are facing in telling animal relationships apart is evolution's signature that some very interesting evolutionary stuff happened here."
It is not as if evolution of new forms of animal life occurred over night, says Carroll, a UW-Madison professor of genetics. The problem is that the resolution of branches that may have taken a few million years to sprout get washed out in the much larger context of 500 million years of animal life on Earth.
"It is hard to distinguish these events, even with boatloads of data," Carroll says. "As you go into deep time, origins are much harder to pick out. And given that there is so much data, you have to ask: why aren't you getting any resolution?"
To illustrate the problem, Carroll notes that if 500 million years from now scientists were to use current molecular techniques to construct the radiation of mammals, they would have difficulty doing so because certain branching events are a "mere" few million years apart.
Using a computer model, Rokas and colleagues simulated just that scenario: What would the mammalian radiation look like had it not happened 100 million years ago, but instead 500-600 million years ago?
"The picture we get is surprisingly similar to the one we get for the animal kingdom," says Rokas. "Certain branches are well resolved, but others looked very bushy indeed."
"As you go into deep time, these bursts of evolutionary origins become harder to resolve," Carroll explains. In addition to the complications of deep time, animal life sometimes has a tendency to explode in radiations as organisms exploit new or newly vacant ecological niches. Famous examples of such radiations include, Darwin's finches in the Galapagos and cichlid fish in African lakes.
In that respect, the results of the new study support paleontological evidence of an explosive radiation at the dawn of animal life.
To arrive at a definitive tree of life for animals, Carroll and Rokas believe, will require much more data and new techniques for extracting the information stored in the DNA record, which will enable scientists to look back in time with greater precision and distinguish the branching events that occurred as new species emerged.
"There are many, many cases where DNA has told us about species' relationships that we never would have guessed based on appearances or other characteristics. We need this tree to understand the great story of animal evolution," Carroll says.