dictate that at the end of every working day our engineers can still
present a working vehicle. It would be an impossible job, and that indi-
cates how remarkable a transition this really was. And now, remarkably,
it is all documented by fossils.
After such public lectures, the question I am asked most commonly is
why cetaceans went into the water. There is much that we still do not
207
208 | Chapter 15
know, but if I step back from all the details and squint my eyes, a blurry
movie reel becomes visible (figure 66). Little raccoon-sized artiodactyls
were eating flowers and leaves, but hid in the water from danger. Their
descendants stayed there, now hiding in the water as predators, spying
for prey. Their descendants learned how to swim fast, pursued new prey,
and little by little, lost the ability to get around on land. After experi-
menting with different ways of swimming, they eventually changed their
bodies to be sleek and streamlined. Thus all ties to the land were broken.
One group added a sound-emission system to its already highly devel-
oped hearing system in order to locate prey: the echolocating odontocetes
or toothed whales. The other group evolved baleen, used for grazing the
krill fields: the mysticetes or baleen whales. There was no single drive to
get from land to water. Cetaceans took small steps, not in a straight line,
and most related somehow to feeding and diet. Each of those steps was
opportunistic, and there were plenty of failed experiments.
In spite of what we know already, there are lots of interesting ques-
tions left, but one strikes me as the Big Question. Mammals, in general,
are highly integrated and built on a more constrained blueprint than
groups such as reptiles and fish. For instance, the dental formula for
placental mammals is 3.1.4.3/3.1.4.3 and hardly ever goes up in num-
bers (see figure 11 and chapter 2). Also, in mammals, there are at most
three phalanges per finger and two for the thumb (figure 13), and there
are around twenty-six vertebrae in front of the sacrum (chapter 12). In
all of those design features, mammals are more constrained than fish,
amphibians, and reptiles. But cetaceans are the exception. They make a
mockery of the mammalian rules, varying wildly in numbers of teeth,
numbers of phalanges, and numbers of presacral vertebrae. It is as if
some very basic mammalian rules governing development have been
broken. And paradoxically, in spite of that greater-than-normal varia-
tion, all modern cetaceans look quite similar on the outside. They all
have a streamlined body, are basically naked, lack a neck, have flippers
for forelimbs, got rid of their hind limbs, and evolved a horizontal fluke
for a tail. If they are built on a relaxed blueprint, why are they so similar
externally? My Big Question is about the genetic switches that caused
the blueprint to relax, and how they affect the paradox of the conserva-
tive external morphology.
The first part of that question is already so broad that it cannot be
phrased as an explicit hypothesis that is testable in the way most sci-
ences operate. Instead, we need to break it up into smaller, more specific
and answerable questions. Some of those answers will come from fossils.
The Way Forward | 209
Only fossils can show us what actually happened in evolution. But many
of the answers will come from studying the genes that channel the devel-
opment of an embryo. Those can only be studied in modern whales.
In our quest for answerable questions, we should start with just one
organ system and make sense of what developmental data and paleon-
tological data have to offer together. Given that feeding evolution is
central to early whale evolution, it makes sense to start there.
tooth development
With it being as difficult as it is to get cetacean embryos, the most
straightforward way to start this project is to go with the dolphin
embryos that John Heyning and Bill Perrin gave me, years ago, when I
was studying hind-limb loss. Those embryos are all from one species,
the pantropical spotted dolphin, Stenella attenuata. As an adult, this
species has more than thirty-five teeth in each upper and lower jaw, way
more than the eleven its Eocene ancestors had. The teeth are tiny, as
shown for a related dolphin in figure 25. More than eleven teeth per half
jaw is called polydonty. Polydonty only occurs in two groups of mam-
mals besides cetaceans: manatees and the giant armadillo Priodontes. In
addition to being polydont, Stenella is also homodont: all its teeth look
the same. There is no distinction between incisors, canine, premolars,
and molars. Homodonty occurs, to some degree, in all modern ceta-
ceans that still have teeth, but is rare in other mammals. Interestingly,
the Eocene cetaceans that I study are neither homodont nor polydont.
Both features show up gradually but more or less simultaneously, start-
ing around thirty-four million years ago. Early baleen whales still have
teeth; there are fifteen to twenty teeth per jaw. Those teeth are more
similar than those of Eocene whales, but there is a definite tendency
toward homodonty. The same happens, independently, in early odon-
tocetes. That makes me think that there is a relation between homo-
donty and polydonty.
We know a lot about tooth development from biomedical studies,
mostly on mice. When the embryo is still tiny, and long before there are
any teeth, a protein is made in the front of the jaw that goes by the acro-
nym BMP4. Another protein, FGF8, is made in the back of the jaw.1
Interestingly, in other vertebrates, BMP4 occurs throughout the jaw, and
FGF8 is not involved in tooth development at this stage.2 And, of course,
those other vertebrates are homodont, or nearly so. Experiments have
been done with mouse embryos, and, sure enough, if the embryo is
210 | Chapter 15
Al igator
Shrew
Mouse
Pig
Dolphin
homodont
heterodont
heterodont
heterodont
homodont
BMP4
BMP4 FGF8
BMP4 FGF8
BMP4 FGF8
BMP4
front
BMP4
back of the jaw
of the
and
jaw
&
nbsp; Baleen Whales
FGF8
Loss of teeth
Homodonty
Homodonty
Polydonty
Eocene Whales
Beginnings of homodonty
Some BMP4 involvement in
the back of the jaw
Heterodonty
FGF8 involvement in tooth
formation in back of jaw
figure 67. Genes determining tooth shape make the proteins BMP4 (black bar)
and FGF8 (white bar). These proteins both occur in the jaw of different vertebrates,
and the pattern varies between reptiles (alligator) and mammals (all others). The
pattern of these proteins in dolphins is different from that of the other mammals.
Alligators and dolphins have similar teeth across the tooth row (homodonty), but
their dental shape results from different gene expression patterns. The branching
diagram at the bottom summarizes evolutionary events leading to the changes in
protein distribution. The Eocene whales of this book are on the segment to baleen
whales and dolphin.
tricked into making BMP4 in the back of the jaw too, the mouse’s
molars become simpler, and all the teeth look like incisors.3 It turns out
that, in dolphin embryos, BMP4 is still present in the front and FGF8 in
the back, but the back of the jaw also has BMP4.4 It appears, then, that
the interaction between these two proteins within the jaw could be part
of an important evolutionary switch: FGF8 taking a role in tooth devel-
opment is a novelty for mammals, and the expansion of BMP4 over-
prints that role in cetaceans (figure 67). Also relevant is that the pres-
ence of these proteins in the embryo actually occurs long before there
are morphological signs of teeth, while the morphological result (homo-
donty or heterodonty) can only be seen when the teeth are formed,
much later in development. That may make it unlikely that those two
patterns, polydonty and homodonty, are the result of one simple genetic
event at a single time in development, although we can’t be sure about
that at this point.
The Way Forward | 211
Our study on dolphin tooth development was a good first step
toward understanding the shapes of teeth and the genes leading to those
shapes.5 It said something about homodonty but did not find a direct
mechanism that linked homodonty to polydonty, and only involved a
single species of toothed whale. I need embryos for more species, spe-
cifically ones not closely related to dolphins, like baleen whales. Those
are even harder to get than dolphin embryos.
baleen as teeth
Baleen whales do not have teeth, but their embryos do.6 Just as with the
hind limb buds, teeth are formed in the jaws of the tiny embryos, and
later in development, these tooth precursors cease to grow, and languish.
These little tooth buds even grow into tiny mineralized structures in
some baleen whales,7 but in no baleen whale do they ever erupt from the
gums. Near the time that the teeth disappear, baleen starts to develop
from the same area of the upper jaw where the teeth used to be.8 The
similarity in timing makes it tempting to speculate that baleen formation
is somehow linked to the cessation of tooth formation. The fossil record
gives some clues, too. Baleen does not fossilize, but it has been suggested
that the presence of baleen in a fossil whale can be deduced from grooves
on the palate.9 These grooves carry blood vessels, and a fast-growing
tissue like baleen needs a lot of blood to supply it. Following this logic,
it has been suggested that some Oligocene mysticetes had the beginnings
of baleen formation even though they still had teeth. In fact, the denti-
tion of those whales was polydont and, to a large extent, homodont.
Similarly, there are many baleen plates, and they are all very similar.
Baleen forms as a thickening of the epithelium of the upper jaw, and
interestingly, teeth initially also form as a thickening. In the case of
teeth, the thickening buries itself into the underlying tissue, the mesen-
chyme. If the two processes are linked, I would expect that a subset of
the genes involved in tooth formation is also involved in baleen forma-
tion. Genes that often work together in building different organs are
referred to as genetic toolkits. It is possible that early in baleen whale
evolution the toolkit that built teeth was reprogrammed to build baleen
instead. That shift of the toolkit caused the teeth to disappear. Such a
novel function for an existing process has been called exaptation. If I
can show that the same genetic toolkit is involved in both tooth loss and
baleen formation, the next question will be whether that toolkit also
operates in other processes—the lack of hair development, for instance,
212 | Chapter 15
since hair development also has similarities with tooth development in
the embryo. If that is true, it is possible that a few changes in a key
group of regulatory genes would affect a whole array of cetacean organs
and drive the evolution of the group.
I think about this as I fly to Barrow, on Alaska’s North Slope. There
I hope to study embryos of the bowhead whale, a species of baleen
whale, that have been harvested by Iñupiat eskimos. Circumarctic indig-
enous people have subsisted on bowhead for many centuries, and the
International Whaling Commission strictly regulates this, so that it does
not affect the health of the bowhead populations, which are growing
right now. I am years away from answering even the simplest questions
related to bowhead development, but it makes me think back to that
first trip to Pakistan, in 1991. I did not go to Pakistan to answer the
questions that I ended up answering, and the outbreak of the war all but
killed my first field season. Only by perseverance and luck was I able to
follow through, and that field season paved the way for the exciting
discoveries that I was part of later. Another decade will be needed to
show how much cetacean embryos can enhance the fossil story.
For now, I am content that this book has summarized the remarkable
progress that has been made in our understanding of whale origins. The
subject used to be undocumented, hard to grasp conceptually, and the
darling of creationists for its absence of fossils
. Now it is the darling of
evolutionary biology textbooks: it is well understood, with plenty of
intermediate fossils, many clear-cut functional links, and the beginnings
of an understanding of the molecular mechanisms that drive it all. Many
questions remain, and without doubt, pieces of this story will have to be
rewritten as we learn more. But that is part of the normal dynamics of
science. New finds are used to test past conclusions, and with every step
we get closer to true understanding. It is also part of the normal dynam-
ics of human life. With every experience that a human has, growth
occurs and old ideas are resculpted. For whale origins, amazing things
have been learned over the past two decades, and I hope a new genera-
tion of budding scientists will push our understanding of whale evolu-
tion beyond our present horizon. It is your turn—go for it.
Notes
chapter 1. a wasted dig
1. R. M. West, “Middle Eocene Large Mammal Assemblage with Tethyan Affin-
ities, Ganda Kas Region, Pakistan,” Journal of Paleontology 54 (1980): 508–33.
2. P. D. Gingerich and D. E. Russell, “Pakicetus inachus, a New Archaeocete
(Mammalia, Cetacea),” Contributions from the Museum of Paleontology,
University of Michigan 25 (1981): 235–46. P. D. Gingerich, N. A. Wells, D. E.
Russell, and S. M. I. Shah, “Origin of Whales in Epicontinental Remnant Seas:
New Evidence from the Early Eocene of Pakistan,” Science 220 (1983): 403–06.
3. D. T. Gish, Evolution: The Challenge of the Fossil Record (El Cajon, CA:
Creation-Life Publishers, 1985).
4. A. Boyden and D. Gemeroy, “The Relative Position of the Cetacea among
Orders of Mammalia as Indicated by Precipitin Tests,” Zoologica 35 (1950):
145–51. M. Goodman, J. Czelusniak, and J. E. Beeber, “Phylogeny of Primates
and Other Eutherian Orders: A Cladistics Analysis Using Amino Acid and
Nucleotide Sequence Data,” Cladistics 1 (1985): 171–85.
5. D. Gish, “When Is a Whale a Whale?” Acts & Facts 23 (1994, No. 4).
The Walking Whales Page 32
