Onto Land and Back: Dr. Maureen O'Leary Studies Whale Evolution

This essay was developed for Week 1 of the AMNH online course Evolution, part of Seminars on Science, a program of online graduate-level professional development courses for K-12 educators.


Why study whales?


Consider these enormous, intelligent animals. They're mammals, but they abandoned dry land over 50 million years ago to recolonize the sea. And they look nothing like the land ancestors they left behind. "They've lost their hair, they've lost their hind limbs completely, and their forelimbs have been transformed into flipper-like structures that look more on the surface like a fish's fin than a forelimb," points out Dr. Maureen O'Leary, a professor of anatomical sciences at Stony Brook University on New York's Long Island. Since paleontologists such as O'Leary have discovered through their research that whales started out as dog- or pig-like animals, they can investigate how this extraordinary transition occurred step-by-step over time. "It's really exciting to study a group of animals that encapsulates so much change, because it's possible to see how evolution has modified organisms in very peculiar ways, and how much they've changed," she points out.

Fossils and DNA: two kinds of data that determine relatedness


Studying whales also happens to be something of a scientific hot potato. It's at the front lines of a debate about the tools that modern evolutionary biologists use to study the history of life.

Biologists used to rely entirely on morphology—the physical features of organisms like the shapes of bones or muscles, or the presence of fins or fur—to figure out the relationships among organisms (their phylogeny, a family tree of species). Comparing similarities and differences among both living and extinct organisms enabled these morphologists to classify them into species, and to construct evolutionary trees. Then, in the 1980s, the widespread use of new tools developed by molecular biologists made it possible to study and sequence the genetic makeup of different organisms. Molecular biologists could now use DNA and other molecules to compare the genomes, or complete sets of genes, of different organisms to unravel their evolutionary histories. The more the genomes overlap, the more closely related the organisms. The availability of these two sets of information—morphological data from extinct and living organisms, and molecular data from living ones—has upset a few apple carts, because the two types of data do not always provide the same result.

How do scientists construct the Tree of Life?


"The best way for scientists to establish relatedness is to use modern phylogenetic methods," says O'Leary. Phylogenetics is the study of evolutionary relatedness among species. This research involves choosing at least three species and identifying heritable features, or characters, to compare across them. For morphologists, these features consist of specific physical characteristics; for molecular biologists, they consist of nucleotide sequences in the DNA of an organism. Both approaches rely on the same computer algorithms to analyze the distribution of those features.
"The scientist codifies those features in what we call a matrix, which looks almost like an Excel file," O'Leary elaborates. "For example, your character might be wing color and your two character states might be blue or red. You look at your organisms, and for every blue wing, you put a "1" and for every red wing a "0." You then simply apply this across as many features as you choose. When you assemble one of these matrices, it's now numerical, so you can submit it to an algorithm that will apply what are called optimality criteria. These are algorithms that generate a tree based on those data. That's how people determine that two species are more closely related to each other than either is to a third species."
Molecular and morphological data have told two different tales of the origin of the whale.

What do we know about whale ancestry?


Whales, dolphins, and porpoises have long been recognized as being more closely related to each other than to other mammals, and so they are united in the group Cetacea. Cetaceans are related to Artiodactyla, a group of mammals that consists of camels, deer, pigs, hippopotamuses, and their living and fossil relatives. These animals typically have an even number of digits on their hands and feet: two or four, unlike the five that humans have. Early fossil whales also had even-numbered digits on their feet, which is one of many features that suggest a relationship to Artiodactyla on the Tree of Life.

whale ancestry

Whale Ancestry
The current understanding of molecular biologists, shown in the cladogram, proposes that whales are closely related to hippos and dolphins. 


As molecular biologists started investigating whale ancestry, they began to find DNA evidence that cetaceans were contained within Artiodactyla, rather than as a sister group to it. This means that the closest relative of whales is a specific artiodactylan—a hippopotamus—rather than Artiodactyla as a whole. In other words, hippos are more closely related to whales than either is to other artiodactylans such as pigs. By the early 90s, molecular biologists were finding more confirmation of this hypothesis while, to their great consternation, paleontologists (or morphologists), were not.

"Not only were scientific ideas changing, but scientific methods as well," O'Leary comments. Skeptical about the value of applying molecular technology to evolutionary questions, some paleontologists were reluctant to believe the molecular evidence. Part of the reason was because the anklebone in artiodactyls is distinctively shaped like a pulley on both ends, and paleontologists have long considered this to be the basis for classifying a mammal as an artiodactylan. "The belief that an organism had to have this ankle to be an artiodactylan was quite ingrained, and many paleontologists were unwilling to consider relationships supported by the molecular evidence until such a fossil was found," O'Leary explains.

Drawing on all the evidence


O'Leary wasn't happy about that. "The concept of having a 'Rosetta stone' character runs contrary to modern phylogenetic methods," she points out. "Paleontologists shouldn't give more weight to particular characters, nor should they assume that certain characters, like a distinctive ankle, cannot reverse. Instead, they should let the data reveal which characters ultimately inform us about phylogeny." She also thought it important to confront the fact that the fossil record contradicted the molecular data. "We can't solve scientific problems by getting rid of evidence," she maintains. "Phylogenetics forces us to back away from assumptions and look at things more baldly, to compare all the data."
There may be good and interesting reasons why fossils and molecules sometimes appear to diverge. Over 99 percent of the organisms that ever lived are extinct, a fact that should give us pause, as O'Leary points out, because living things are just a snapshot of life on Earth. "We don't have fossils of all of that 99 percent either, but we do have a lot of fossils, and they do tell us a lot of the actual history. Reconstructing the history of life using less than 1 percent of the available data from living things alone may lead us astray."

A fossil find resolves a dispute—but leaves other questions unanswered


Both the fossil and the molecular record have their advantages and disadvantages, but each records the same story. Since the late 1970s, University of Michigan professor Philip D. Gingerich has been searching for evidence that would resolve the whale-evolution debate. Backbones were abundant, but hands and feet missing. Finally, in Pakistan in 2000, Gingerich discovered the 47-million-year-old fossil whales Artiocetus andRodhocetus, the latter with a fully developed hind limb and both with ankles very much like the artiodactyls', down to the double pulley. "That's gone a long way to convincing many paleontologists that whales and artiodactyls are close relatives. It just took us a while to find the fossil," says O'Leary.

Combined analyses of molecules and anatomical information produce phylogenetic trees that indicate that whales, fossil and living, are artiodactyls because they're consistent with trees based on molecular data alone. So scientists have now replaced the term Artiodactyla with the term Cetartiodactyla to describe the common ancestor of whales and artiodactylans. Still a mystery is the fact that the newly discovered ankle does nothing to reinforce whales' link to hippos. "It fits the artiodactyl group, but there's nothing that makes it hippo-like as opposed to pig-like or camel-like," O'Leary elaborates.

Looking for more "walking whales"


Further finds may fill in the puzzle. Recent excavations of intermediate fossils (between terrestrial and marine life forms) in India, Pakistan, and Egypt have sparked increasing interest in whale paleontology. "For example, they've found an animal not much bigger than large dog called Ambulocetus—that's Latin for 'walking' plus 'whale'—with large legs that look like they could support its weight on land. Someone without knowledge of its evolutionary history might say it looks like a crocodile or a dog, but we can tie it to whales phylogenetically because of certain features of the teeth and ear region," O'Leary recounts.

O'Leary works mostly in the Republic of Mali in West Africa. The northern part of Mali is part of the Sahara Desert, but Mali was once inundated by a shallow sea that ran north to south and cut North Africa in two halves. "This means that there are now exposed fossils and rocks of marine life from about 55 million years ago, early in placental mammal evolution, in the early part of what we call the Tertiary period. It's my hope that we will ultimately find whale fossils in this area, the way paleontologists have elsewhere in Africa," she says.

Whales have more to teach us


Why did a group of terrestrial mammals abandon life on land for life in the sea? The answer, as scientists piece it together, has much to tell us about the pattern and process behind a major evolutionary transition. "If you can reliably, in an evidence-based fashion, place whales within the context of mammals like sheep and hippos, it's hard not to step back and say, 'Wow, it's amazing that evolution is capable of transforming an organism that much over 50 million years or less,'" O'Leary points out. "You're looking in the broadest sense at change through time, and in those terms, whales are where it's at."


Nature: Walking with Whales (PDF File)

A summary of the fossil evidence that helps fill in the gaps in understanding whale evolution.

Berkeley: Introduction to Cetaceans
A brief article about whales and dolphins and their developmental history.

Stony Brook: Vertebrate Fossil Laboratory
Get an inside view of Dr. O'Leary's research at Stony Brook University.

Discovery: Walking with Prehistoric Beasts
Explore some of the ancestors to several modern species, including Basilosaurus and Ambulocetus, both closely related to whales.

©2007 American Musuem of Natural History. All rights reserved.