Evolution and Phylogenetics Expert Interview Transcript: Philip D. Gingerich, Ph.D.

Gingerich is a professor of geological sciences and director of the Museum of Paleontology at the University of Michigan. He has studied the evolution of archaic whales for over twenty-five years, collecting specimens in Pakistan and Egypt. In 2000, he found fossils that confirmed the assertion by molecular biologists that whales evolved not from mesonychids, extinct wolf-like animals, but from artiodactyls, the ancestors of hippos and camels.

How would you define phylogeny?
Phylogeny comes from — entomologically, it means leaf — and so it’s a metaphor referring to the venation in a leaf, and the idea that life started with a single stem, and that has branched into all the branches that you see in the venation in a leaf. So it’s to give a sense of a tree-like or leaf-like structure of the history.

How is it used in evolutionary terms?
Well, in evolutionary science the stem is the origin of life, or the particular origin that a particular scientist is interested in. I’m interested in whales, and so for me, this stem is the early things that whales came from, and then the branches of the venation are the subsequent branching off through time. Now, of course, a vein — the venation of a leaf starts at an apex and branches from there and that’s the way life goes, through time, so it gives an arrow or a direction to time — in the direction of more branches.

What is phylogenic reconstruction?
Phylogenic reconstruction is trying to put together a picture, a leaf-like picture, a venation-like picture that shows how the animals living today and the animals that lived in the past are connected to each other. If they’re ancestors, they would be on the same vein and if they’re what we call “sister taxa” they would be at the terminals of different branches.

What is a cladogram?

A phylogenetic tree is the history of life through time. And a cladogram is a branching diagram, usually shown as a kind of a cone-like structure, that shows how the animals are related to each other without reference to time. There’s an implication that things that are closer together on the cone are separated by less time but that’s not explicit.

Can you talk a bit about the evolution of the field of paleontology itself?
Well, we start with observations. We start with what we can see, and in the study of evolution what we saw at the beginning, in the 1830s, ’40s, ’50s, were fossils recorded from geological strata in Europe starting in England and in France and the sequences were the same-the fossils were buried in the same chronological pattern-and then that was generalized across Europe and around the rest of the world. And it’s still true that the sequence of fossils found in geological strata is the same everywhere-not exactly the same but the general pattern is the same.

Now the geological strata have to be older on the bottom and younger on the top. You can’t put a stratum on top of another one if the first one isn’t already there so that tells us the ordering in time. And in the 1840s to 1850s, this gave rise to the geological time scale, so the geological time scale is about the Cenozoic, the Mesozoic, and Paleozoic ages-the “zoic” refers to zoo or animal life-so the geological time scale is based on this sequence observed everywhere through geological time.

Now it didn’t take very long once that became established to realize that since the animals are different they must have been changing through time and so that very same set of observations is the basis of our understanding of the hypothesis of evolution and I would say it’s been a hypothesis so long that that part of evolution is an established fact of geology, the sequence of life through time.

So why do we still call it the Evolutionary Theory?
Now, scientists are reluctant to define things as true. We don’t pretend to know the truth, and so we always speak as if we may learn something yet that we didn’t know that might change what we think and that’s why this is all called a theory.

In terms of evolution, the process of how it works and all the possibilities of how it can work are so many and so complicated and still so poorly understood that it’s really appropriate there to be very careful when talk about hypotheses, and we try to test them, and that’s how we proceed in our science.

What is a synapomorphy?
A synapomorphy is a technical term that is used when you know that two animals have inherited the same characteristic. First, they have the same characteristic and you know that they inherited it together. Then you have a synapomorphy, as opposed to two animals having the same characteristic that they inherited independently, and that would be a symplesiomorphy. So these are technical terms to try to distinguish the possibility that things can arise from different ancestry.

How could you use those traits to construct a phylogenetic tree?
Well, here’s where we cladists and paleontologists diverge a little. To know that you have a synapomorphy, you have to know the phylogeny. So logically it’s difficult to understand how you can use synapomorphies to reconstruct phylogeny.

And what happens in practice is people see characteristics that are shared, and if they’re special enough or unique characteristics, we assume that they’re synapomorphies, and then we group together the animals that have more of these synapomorphies and leave aside the ones that lack them or have fewer of them and in that way we can rank taxa farther and farther away from the core, and it is commonly assumed then that that’s some reflection of the phylogenic history. It’s a kind of an illogical central dogma is what it is. It’s taken over in the last 20 years and I was educated before it, and so I don’t understand it, the logic.

Because?
Because to know that something is a synapomorphy, you have to know the phylogeny. So how can you then use that to reconstruct the phylogeny? Now it’s done with computers that maximize the parsimony, and thus minimize the number of changes so there’s a certain logic there.

How about homoplasy?
Homoplasy is the idea that the same characteristic can arrive by different pathways and not reflect the ancestry, so the alternatives to synapomorphies.

What are the main methods used in classification history?
Classification traditionally has been about putting things together that look alike. In our refrigerators, we put the eggs in one place and we put the milk in another place, generally, to organize it, and we might have skim milk and whole milk but we put the milks together. We might have brown eggs and white eggs but we put them together. And so, in our daily lives and in biological science, we put things together that seem to be alike.

Now, the main thing that we’ve learned, as we learn more about evolution, is that things that look alike can come from different sources. So, in Australia, we have a marsupial mole that burrows in the earth and here in North America we have a placental mole that burrows in the earth. They look very similar and have completely different ancestries.

So, while you might at a first attempt put all the moles together, when you study them more carefully you realize, “oh, these are completely different in terms of their reproduction. Geographically they’re completely different.” And so we separate them now and we don’t put them together because we realize that they have a different ancestry.

So that’s the main thing that has changed. As we have learned more and more about the history of life, we’ve had to separate things — make two categories of moles or make a category of marsupial wolves and placental wolves and there are many such examples at a high level in the classification and then also at a low level. And in many cases, we don’t know how things are related exactly at the very fine scale so there’s a lot of uncertainty at that level.

What are the different levels of data?
If we go back to the mole case, we started off with the anatomy. Any of the major divisions of the classification of animals–and indeed of mammals–are based on their reproductive structures, because these are very conservative in evolution, we think.

And so initially, studying animals anatomically, they were sorted into groups. Then as we learned more about the history of them through the fossil record we were able to add some of this historical phylogeny to understand the differences.

Now in the last 20 years or so, with the distribution of computers on everyone’s desk, we’re able to take many, many very fine anatomical characteristics and process these and do a what’s called a “parsimony analysis” and the “cladistic analysis,” and really sort out the anatomical details and compare them across many taxa in a way that would be very difficult to do by hand. And then maybe the most recent development is sequencing DNA and the beauty of that is that it gives us many, many characters to study.

The difficulty with DNA analysis is that the characters are very simple, they’re all just letters, and there are only four of them, and so the problem I mentioned before, the complexity of homoplasy is very great because these nucleotides are constantly mutating and writing over and expanding the genetic code.

What were Hennig’s contributions to phylogenetics?
Willi Hennig was a German entomologist studying insects, which are very complicated, and it’s very difficult to know how they’re related. And in about 1949 or ’50, he published a book in German about what he called “phylogenetic systematics” and this was the idea that you have to take phylogeny into consideration when classifying things, and this is a big problem in insects in particular because it’s so difficult to recognize phylogeny. It’s important when you study mammals, where it’s more clear, but obviously it needs to be taken into account in insects too.

And so he published this book. It was translated into English maybe in the 1960s-late ’60s I suppose- and at the same time computers were becoming available, and so this fit right into the idea that we should break things down into their smallest characteristics and then analyze these by comparing them with each other to find the most parsimonious possible history we could use to explain the distribution of characteristics.

What are the rules of phylogeny classification?
To me there are three, and these are, one, superposition of geological strata giving us an order through time, and two, faunal succession, the observation that the faunas change through that giving us a sense of their evolution and dividing up into classes over geological time.

And then I would say that the third principal is the idea of intermediacy, or continuity from sample to sample, as the evidence that they are related so that we wind up tracing lines up through geological strata, often in very broad brushstrokes in the sense of tracing mammals as a group through the geological strata from their origin in the Triassic and the Mesozoic to the big diversification at the beginning of the Cenozoic and up to their abundance today.

Or tracing smaller lines within a group, like horses and other odd-toed perrisodactyls up through time to understand their history. So the intermediates are important evidence of transitions.

Does this hold up at any level of taxa?
Yes, it does, but of course, at the finer levels, it’s more difficult to find and know the connections. Now, this is also limited because there’s a vast array of life, especially insects, and there are other groups that are equally diverse and are almost as diverse and also underrepresented in the fossil record where you can’t trace their history through time. So if you want to classify them rationally, you have to do something else, and this is really where Hennig’s phylogenetic systematics have come in, to give scientists who don’t have a fossil record a way to approach the classification in a parsimonious way.

Can you talk about your own research interests?
My interest in whales is one of those happy accidents of science that you look back at and you realize was just good luck. I’ve worked since the beginning of my career to try to understand where the modern orders of mammals come from. We can trace horses back in time to the beginning of the Eocene. We can trace primates — the group that we belong to-back to the beginning of the Eocene. We can trace the split-hoofed artiodactyls back to the beginning of the Eocene.

And my fieldwork, initially in Wyoming, was to try to take those back farther to see where their ancestors are, and very quickly we established without any doubt that they can be traced back to a particular bed and they are never present before that. So they must have come from somewhere else, and to try to figure out from where, I started a project in south Asia because I thought, “this is a time when generally climates are warming. Probably things are coming from the southern parts of continents into the northern parts of continents.” It’s very easy to imagine, if it got warm enough, that the continents would be connected at high latitudes — that’s where the land bridges are-so it has to be warm to cross them.

So, I started a project in Pakistan to try to find — to try to get a south Asian perspective on what I was seeing in Wyoming.

Well, I went there looking, following a very slender lead, which was a fragment of a jaw with the teeth broken out of it. It was said to be a land mammal, and when I got there I saw that it was completely marine. We found some hipbones and things that clearly held a big hind leg and we laughed because we thought, oh, these could be walking whales but it was too preposterous to believe and so we ran away, literally. We ran to a different part of the country. We drove to a different part of the country and started work there and very luckily found a tiny jaw at the end of the first field season, so I kept going.

The next year, one of my colleagues found a skull in a block of rock. We couldn’t see much of it but we could see that it had a brain about the size of a walnut and it was a skull bigger than a wolf. We brought it home, cleaned it, looked at the ears, and the morphology of them and they looked like the ear bones of whales.

And suddenly I thought, well, this helps explain the walnut-size brain because whales have big skulls and relatively small brains-some of them do today, and they would have in the past, the archaic whales would. And we were near a marine setting near the shore, so this was plausible. And the more I thought about whales after describing that find, the more I thought about them, the more I realized this is a big transition. This is the origin of a modern order of mammals, yes, but it’s not just horses coming from other land mammals. It’s not just primates coming from other land mammals.

This was something that had come from other earlier mammals, but at the same time, it was going back into the sea. So it’s a big evolutionary transition backward. It’s fascinating for a person interested in evolution, and so ever since then for the last 20 years I’ve been really consumed by the idea that we have to find all of the links, all of the transitional forms from whales living today back to the transition from land animals back into marine mammals, and that’s what I’ve been working on especially on the early part, the archaic whales from the Eocene.

What did earlier discoveries tell us about whale evolution?
Again by comparing their anatomy, we’ve known that whales are mammals. They’re warm blooded. They have occasional hairs. They don’t have much hair. They nurse their young like mammals do. That’s the definition of a mammal. And so we’ve long known that they belong with animals that are predominantly living on land today. Mammals have a very good fossil record going back 200 million years, and it’s virtually all on land.

So just from that evidence without any direct evidence from fossils we’ve long thought that whales evolved from a land animal and must have gone into the sea at some time. But of course, it’s one thing to hypothesize this. It’s another to find evidence of the steps by which it happened. So that’s what we’re pursuing.

Did you start with opposing theories of the evolution of whales?
There were no opposing theories. There were no real theories. There were no real hypotheses because whales are so different, that nobody knew what to relate them to. And a very influential classification of mammals by George Gaylord Simpson in the 1940s said this explicitly. He said, “I don’t know where to put whales. I’m sticking them here but I don’t have any reason for it.”

Now a new science of molecular biology was just being born then and two other scientists involved in this in New York City saw what Simpson said about whales and thought, I think we can do something about this. And so they reacted the blood of whales with the blood of all the other major groups of mammals, to see how much immune reaction they would get. If things are distantly related, generally you don’t get much reaction, and if they’re closely related then you do.

And so the blood from all the orders of mammals showed no immunological reaction except one, and that was the split-hoofed ungulates; the split-hoofed artiodactyls showed a big reaction. So they said, “based on this immune reaction, we think artiodactyls are the sister group or the closest living group to whales. Now at first, that seems preposterous, because the hoofed mammals eat plants. Whales eat animals. They live on land; whales live in the sea. And there’s not any good anatomical reason to even connect them.

But then a scientist at the University of Chicago, Leigh vanValen, saw a way to tie whales and artiodactyls together in a group called “mesonychidae.” These are part of a group called condylarthra. These are archaic Paleocene mammals that survived into the Eocene. They have teeth that look like archaic whales and feet that look a little bit like artiodactyls. They have hooves, is the main thing they have. They still have basically a five-fingered hand and foot but they have hooves.

So he proposed, “okay, we can connect whales to artiodactyls, like the molecular biologists would like to, through mesonychidae.” And it turned out that the mammal that I went to Pakistan to find, in search of more mammals, that fossil that I went to see had been described as a mesonychid, this group at the base. It turns out it was a kind of whale, and some other things that were described as mesonychids in Pakistan turned out to be whales.

Well, as evidence in support of vanValen’s idea, several of the mammals in Pakistan known from their teeth that were described in the 1840s-1860s as being mesonychids have turned out to be whales. The teeth can be confused, in other words, between the two.

And I would say that we, as paleontologists, have all been happy with this idea that it resolves the connection proposed by the immunologic results with the fossil record as we know it, and we expected that mesonychids were probably the connection between artiodactyls and whales.

So you looked for that connection in Pakistan?
This idea of vanValen’s about mesonychids being at the base of whales and artiodactyls has been a good one, but we’ve long known that what would cement it would be finding the ankle bones of early whales. There are two parts of mammals that are most informative about what it’s related to. One is the teeth, and the second is the ankle bones because they’re complicated, they preserve well as fossils, they’re dense bones, they bear a lot of weight; they’re good fossils.

If we can find those, we could corroborate, I thought, the idea that archaic whales evolved from mesonychids. Let’s just go back a little bit. In the 1970s, I got interested in whales because we found the one that we named Pakicetus, and then shortly afterward, the Soviet Union invaded Afghanistan. It wasn’t possible to work along the Afghan border anymore, and I had to find another place to work.

So during the 1980s I worked in Egypt and worked in a fantastic place there-now a national park-called valley of the whales or Zeuglodon Valley. And we found hundreds of whales there in the desert, and we started to look for parts of them that weren’t well known. How many vertebrae they had wasn’t well known; it wasn’t known at all what the hands looked like, and of course, we thought the feet were gone. We thought that they, too, would be reduced like they are in modern whales.

At the end of six seasons, we found the feet of whales, and we found that they still had toes. The only problem with that was that the ankle bones themselves, which would have been most informative for comparing them to other mammals, were too modified to be able to compare them. But I thought, if we have feet on whales ten million years after Pakicetus, then we can find the intermediates. In other words, ten million years is a long time, even to a geologist. That’s a long time and we have a good chance to find the things in-between.

So I started in the 1990s, a project in Pakistan, looking for these. And finally in 2000, finally, we found it. We found many skeletons with the hands and the feet and the ends of the tails pulled off by predators and taken away. And finally in the year 2000, we found whales that were in water deep enough or shallow enough — I don’t know exactly — that the predators didn’t steal the hands and feet and tails. And we found that the ankles of whales are not like mesonychids. And I can tell you that for the first week after finding this in the field I was in denial because I couldn’t believe that the ankle bones are exactly like those of the artiodactyls. I couldn’t believe it. So mesonychidae are out of the picture, and whales and artiodactyls connect to each other. Mesonychidae is still there somewhere on the side but they’re not part of this story.

It only took a week to get over the shock, because we found that first find was two bones. One called the astralagus, that’s the main hinge bone in our ankle, and the other was the bone called the cuboid that sits just on the side of it in front.

And the astralagus in artiodactyls is completely distinctive. It has a double pulley, one on each end. Kids play with them as cars around the world. They’re commonly used as toys because they look like they have four little wheels on them. And those are the two hinge joints on each end. And you could imagine that this evolved in parallel, that this was a similarity but not a synapomorphy. A similarity, but a sympleisomorphy, so it wasn’t meaningful.

But the cuboid, I looked at it and it was very distinctive and it had a big notch in the side of it. After a week I was able to find a goat ankle in the field. We were collecting fossils. We didn’t have things there to compare to, but you do find like bones of other animals sometimes, living animals, and we found a nice goat ankle. And it has a perfectly notched cuboid, and I realized, “get used to it. This is — this is going to be the answer.

And then when we got home we had collected several hands and feet in the field in plaster jackets, which is the best way to do it so you don’t disturb anything, you don’t lose anything. You bring it home and you prepare it in the lab. And so we got those home, and we had not only the astralagus and the distinctively notched cuboid, but also the calcaneum, which is the bone that makes your heel. It has a very distinctive curved surface on it for the fibula bone to articulate with, and that was there, too, so it’s over. There’s no doubt about it anymore.

What response did you expect from the community?
An instant one, and I knew this would be true, so we had to work quickly to study this material to be able to publish it while trying not to let word out about what we’d found, because I knew that people would embrace this idea immediately, and it’s good to show the evidence before you leak the word out.

How did the molecular data corroborate your find?
I was pretty sure that this idea of whales and artiodactyls being sister groups would be accepted immediately, partly because our evidence was so good, ankle bones associated with complete skeletons, virtually, and partly because our molecular colleagues have been increasingly strident that whales and artiodactyls have to be closely related. And they even think that they can see where whales came out of the artiodactyls, and they propose that hippos are the ancestors–or are the group that’s closest to the ancestry–of whales. And our evidence has basically come out supporting that.

I had read what they wrote. I was fully aware of it, but I never fully believed it because the molecular people can’t see groups that are extinct, and it would be very easy for mesonychids to be in the middle and just not to be able to see them just by looking at living animals. But in this case, now the fossil evidence supports a connection of whales and artiodactyls without any mesonychids at the base.

Is it rare to unify the molecular data with fossil finds?
This is a very rare occurrence. This is virtually unknown. We have 18 orders of mammals living today: the horses, the artiodactyls, the whales, the rodents, the rabbits-18 of these groups. They can, most of them, be traced back to the beginning of the Eocene, and then we mostly don’t know how they’re connected.

And so our molecular colleagues are telling us many answers, some of which disagree with each other, none of which we can fully believe because we don’t have evidence from fossils that tie them together. So this is a very special case to be able to tie the groups together with fossils too. It’s what we’ve been working, working, working to get for all these groups, and at least now we’ve got it for one.

Were you surprised that it should happen with whales?

DNA is not an ideal evolutionary document and the reason is partly because the code is too simple. It’s very long, so it gives us lots of information, but the letters coded in it are too few to convey a lot of information, so it’s easily written over.

How can we use DNA as an evolutionary document?
The second problem with it in terms of the history of life is that you can only study it in things living today. It doesn’t have a fossil record. It doesn’t have something you can trace back and see the steps. It just has connections and you have to jump from one to the other. There are no intermediates that let you trace the lineage, and so that’s the main difference. With the fossil record the game is, the objective is to trace all the lines.

Now, having said that, the problem with the fossil record is that many groups don’t fossilize well so that the overall objective should be to use the two lines of evidence together. Use one, the fossil record, to give us the historical perspective for the things where we can learn that, and then use DNA evidence to learn how to make inferences about the living things that don’t have fossil record.

Can molecular data help with evolution theory?
I don’t think molecular data help prove that life has a history that we call evolution. I think molecular data corroborate it. I think they have to be interpreted in light of that history, but I don’t think they add a lot, and the reason is just because it can’t be traced back in time, and if you’re talking about evolution as a history of life you have to see some steps through time.

So it doesn’t help us with that, but it certainly mostly corroborates what we see, and in one case where it argued against what we saw this whale-hippo connection turned out to be plausible so it may give us more detail in places where we have uncertainty from fossils.

Can we work backward from one organism to construct a phylogeny?
Paleontologists work backwards from the organisms living today. For example, if I show you an Eocene horse at the beginning of horse evolution, I doubt that you would know that it was really a horse. The teeth certainly look nothing like modern horse teeth. The ankles do, so that might convince you. But as far as the teeth go, you can only appreciate that they’re related by seeing what’s present in the geological column back through Cenozoic time, back to the beginning of the Eocene.

And you see all these strange and wonderful things, very high crowned cheek teeth of horses, very complicated crown pattern. You see all of that developing out of very simple molars just like yours or mine when you trace them back through time. So in that sense that’s how we work back from what’s living today, geological bed by bed, backwards through time, to see what the history was of the animals evolving forward through time.

Can you address the metaphor of phylogenist as mapmaker?
I would say that studying phylogeny is a lot like map-making was when you could only see very small pieces of the whole picture at any one time. And of course we’ll never get to the point where we can have a time machine that’s like a satellite giving us the whole picture like a map maker has gotten today. But it’s very much like working and making maps in the early days.

Phylogenetics is a necessary part of biology; are evolution and biology becoming one?
Theodosius Dobzhansky is famous for the claim that nothing in biology makes sense without evolution, and by that, he means evolution in the geological sense, evolution through time-he means phylogeny. And it’s true that to organize the information in biology, which is overwhelming — I mean, if you have any idea how many organisms are living and how complicated they are, a person is easily overwhelmed. Without some way to organize that, to simplify it, to be able to talk about groups of animals in a meaningful way, it’s overwhelming, and nothing can make sense without understanding the basic phylogeny.

Now, some of that we know today, but a lot of it we’re still working on. Especially the deep roots, how they’re related to each other, and then many of the details up higher in the phylogenic tree are still uncertain as well.

So is phylogenetics the best method available?
For many groups, what is called phylogenetics and even molecular phylogenetics is really the only way to approach how they might be grouped together, how in a logical way they might be related to each other. The difficulty is that in those very groups, we have nothing to test the result against and so it’s not really possible to know whether we’re getting right answers or wrong answers but at least it’s a way of doing something.

What are the main driving questions?
There are always questions — there’s always information missing and for a paleontologist interested to fill in all the gaps, you’re in a mad chase, because every time you fill a gap you’ve made a new one above it and a new one below it so you’ve doubled the number of gaps. Every discovery makes more gaps. But the gaps get smaller and so in that sense, they’re satisfying.

One of the things I’m very interested to know is what the tails of archaic whales looked like, because we don’t have them yet. We don’t have them from the early ones. The one mounted up behind us [points to skeleton suspended from ceiling] has a tail and clearly was fluked and tail-powered — the locomotion was tail-powered like whales are today.

Somewhere we went from hind limb locomotion to tail-powered locomotion in whale evolution, and we don’t know when. To know when, we’ll have to find good tails so that’s one thing I’m pursuing.

What do you want teachers to understand?
Well, I think one of the most important things for teachers to appreciate about paleontology is that it is a search for sequences of fossils through time. It’s a search for lines of evolution through time, the evidence of which comes from the past. We use historical evidence to study a historical problem. We’re not making it up from the present. We’re actually documenting step by step what happened at different times in the past.

I would say that the thing I would most like biology teachers to understand about the work that I as a paleontologist do and other paleontologists do is that in studying fossils we’re studying animals that lived in the past. When we have sequences of fossils up through time we’re studying lineages and how they change through time. We’re not making it up from the present. We’re studying it when it happened and that’s the most important thing about what paleontology contributes.

Can you explain this diagram? (refers to ‘Revised Evolutionary History of Whales’ figure.
Here’s an example of what we mean by a phylogeny. And normally we put time on the vertical axis with the idea being we’re coming up through time. Here I have the end of the Mesozoic and all of the Cenozoic, all of the recent animals’ life, starting about 55 million years ago. That’s what we call Paleocene. Then we have the epoch called Eocene to 35 million years ago, and then up to the present.

I’ve added a dashed line here just to emphasize that 65 million years ago is the extinction of dinosaurs, and the time when mammals diversified, so that’s what this represents, just for one group of mammals but we can represent others this way too. The group that’s represented here in this bubble, as it were, are the archaic hoofed mammals-what we call condylarthra -and the mesonychids are a part of that.

Over here coming back from the present-all the way back to the beginning of the Eocene-we have the artiodactyls, the split hoofed ungulates; split hoofed mammals. And here in the middle, because we’re emphasizing whales, we got the whales coming back almost to the beginning of the Eocene. We don’t have them all the way to the beginning yet, but we might in the future.

And I’ve divided them into the two groups that are living today, the odontocetis, the tooth whales, and the mysticetis, the baleen whales. And they come back, not to the beginning of the Eocene, but they come back to the beginning of the Oligocene epoch, which is a time when the earth’s climate became reorganized for some reason. Cold was concentrated at the poles. We have the climate we’re familiar with today starting at 35 million years ago, and that’s when modern whales diversified. Before that, all of the whales belonged to a group we call “archaeoceti,” or archaic whales.

Now, the question about whale evolution has been tracing modern whales back to archaic whales. Do they then connect to the mesonychid condylarthra, or, as our molecular colleagues have argued, do they connect directly to the artiodactyls?

As paleontologists, we’ve thought they go through a mesonychid connection. Artiodactyls would come out of that and whales would come out of it, or come out of condylarthra at least, and the connection would be early.

Now we think, because of finding these astralagus bones and other evidence, that the artiodactyls and the whales are more closely connected to each other and the condylarthra are outside of that connection.

What are the techniques for illustrating this type of lineage?
We’ve given some shape to the “balloons” here. They come back and connect. That indicates that we think there’s a relationship there. I’ve left some space between the balloons, the background color, just to emphasize that these are different groups separated from each other. So in some sense we’re seeing the morphology, the shapes of the animals, on this horizontal axis, not really quantified, and this vertical axis is time quantified in millions of years.

Can we visualize evolutionary patterns?
This is the way that we represent the pattern of evolution graphically. It’s like drawing the positions of the continents on a map, and we plot something about the morphology against time and then in that space we array what we’re interested in.

Could you represent the finer level of organisms?
This map can be drawn at any level of detail. I’ve drawn here a very distant picture, a very broad picture. I’ve divided the whales into the two suborders and some of the common names of the members are representative here and at the archaic suborder. In the case of the artiodactyls, I’ve shown the principal families that are represented here, or at least some common, familiar representatives of those.

Hippos, for example, could be traced back into the Oligocene, possibly as far as the middle Eocene, and the others have relationships that we could show by their connections, but I haven’t shown it here. Now another thing to emphasize, though, is that a reason for laying out the evidence like this is to explain that the evidence that whales living today evolved from artiodactyls doesn’t just come by jumping from one to the other. It comes by connecting them back through time with many intermediates and the strength of the evidence depends on how many and how complete the intermediates are.

So that you could expect me to lay out on a table the representative skeletons of the whales that are known that connect modern whales back to artiodactyls. Now it would be a big table and we’d have to go around to museums around the world to find some these because they aren’t all in our museum here. But that’s what we consider evidence of an evolutionary connection.

Where is fossil found in 2000 on that chart?
It’s right down near the base. It’s not all the way at the base. The fossils we found in the year 2000 are illustrated here, and in time they would fall right about here, so the earlier ones are not so well known. The connections here are not so well known and this is an emphasis of future research is to find intermediates right at this node at this connecting point.

The missing pieces of this puzzle are right down here still. There are some other smaller pieces missing throughout but the important pieces that are still missing are these right down here.

Unexpected piece would result in having to redesign?
If we find something that doesn’t fit the puzzle, we have to rethink it, and a good example of that is, I would never have drawn this connection here before. I would have drawn the green connecting into the condylarthra, and then the condylarthra connecting over to the artiodactyls and so this is a piece here-this connection-this is a hypothetical piece because we haven’t actually found the evidence yet but this whole hypothesis is new since the finds of 2000.

Are there morphological trait examples?
I’ve illustrated the astralagus bone, that’s the pulley bone in your heel; in your ankle that lets you raise and lower your foot-it’s a hinged joint-and that hinges this one. Artiodactyls have a second hinge at the base of it here as do early whales and these being so similar is evidence that the two groups belong together.

If, as I originally thought, these evolved through an intermediate here, we expected to find an ankle bone like this, but that’s not what we found. Instead we found something that’s much more artiodactyle-like and that’s what leads us to hypothesize that this is the connection.

If we had done this based on teeth, which is how we did it before, the connection would have been through here. Now, in emphasizing the ankle bones over the teeth we have to explain the teeth away as a homoplasy, as convergence, and our feeling is that the complexity and the detail of this particular similarity is greater than the complexity and detail of the tooth resemblances that we emphasized before.

Now, remember, this is now a new hypothesis for us. We’ve rejected the old one. Whether this holds up depends on what we find when we find the missing pieces here. But we are now predicting that all of these will have this kind of ankle. If they don’t, we’ll be very surprised.

Is this also the most parsimonious possible diagram?
This is the most parsimonious considered back through time.

Have methods changed over time?
It’s important to understand that the word “phylogeny” is used in two senses. One is the traditional sense going back to the etymology of the word, going back to the last century, to the 19th century. I mean of a branching diagram starting from a stem with the branches separating up through time. Time is an axis of a phylogeny.

Now the word is also used by people doing branch of this called “cladistics” where time is not involved and they are simply taking the characteristics, drawing a branching pattern here, and saying that this is outside these because these are more similar to each other. So this diagram could be simplified and represented as a branching pattern. The thing that would be lost would be the dimension of time and the whole idea of evolution is one about time, and so this is why we continue to put the axis here.

How can computers aid in finding the relationships?
Now, this is very simple to say that these two are more similar to each other than they are to this one. But if you have three or four, ten or 20 or a hundred or a thousand of such comparisons, some of which are conflicting-I’ve already told you that the teeth conflict with this pattern-how do you evaluate those? And that’s where computers come into it, and we have computerized cladistics.

Do you use computers in your research?
No. I mean my students do and I don’t, because conceptually, this is how I think about the problem and so this is how I represent it.

What are these fossils?
These are the hand on the left and the foot on the right of Rodhocetus, which is one of the new finds in 2000. And a skull of Artiocetus which is the other one we found in 2000 that has a skeleton with ankle bones preserved in it. The best hands and feet are Rodhocetus. The best skull is Artiocetus, and so that’s why I show them here.

The hand is interesting mostly because the digits of the hand terminate in little hooflets. The foot is interesting mostly because it’s very elongated just like a flipper you’d wear if you were scuba diving or swimming at the pool.

And instead of hooves, because these digits weren’t weight bearing, they have elongated, more pointed terminal bones that would be inside the flipper. And then the ankle bones are preserved here in a way that’s difficult to see so I’ve got casts of them on the side here and we can point out the double pulley on the astralagus, the fibular facet here and the notched cuboid which are characteristics that are just like those in a pronghorn and artiodactyls, the double-pulley astralagus, the fibular facet, the notched cuboid here are very similar to what we see in Rodhocetus.

Other characteristics that are interesting in these archaic whales if we look at the skull are the length of the rostrum, the opening of the nares, the nose, are very far forward on the skull so they haven’t retreated up the skull yet like modern whales do. And the breadth of the frontal shield here, as we call it, is relatively narrow. The teeth are present in the full number and the differentiation into incisors, canine, premolar teeth, and molar teeth that we’re used to seeing in virtually all living mammals.

And I can mention also the ear bones here, the auditory bulla, is very dense and large, and that’s like whales that have to hear in water, so these are certainly partially aquatic.

What is the significance of these characteristics?
Well, the characteristics that you see in the skull here are similar to those of later archaic whales, and make a transition into modern whales.

Modern toothed whales — the earliest of the modern toothed whales and baleen whales still have these complicated teeth like archaic whales do, so they carry the characteristics forward in time.

In the ankle structure, the characteristics that are most interesting that are the ones that I mentioned here, the double-pulley astralagus, the fibular facet, the notched cuboid that are distinct from what mesonychids have and are similar to what artiodactyls have that leads us to associate them with the artiodactyls instead of the mesonychids.

Now, the idea of what is a characteristic of interest, what’s important, what’s a character in a character-based analysis depends somewhat on the investigator, because someone might code the pulley on this end as one character and the pulley on this end as one character. Someone else might think, oh, they’re related so we’ll just code that as a single character, double-pulley astralagus and similarly characters can be divided or subdivided or lumped together by different investigators depending on what they think their function of significance is. So that’s a source of some uncertainty and controversy generally in this kind of study.