raoellids are from the Chorgali Formation of Pakistan, around fifty-two
million years ago, and the youngest are probably from Kalakot, approx-
imately forty-six million years old. The family is not a very coherent
group. Scientists have referred new fossil specimens to the group without
studying all those that were already known; as a result, the group has
become a somewhat chaotic assemblage. It would be useful for someone
to study the entire group carefully. But such a systematic revision will not
be easy. Many species are known only from a few teeth, and the fossils
are dispersed over three continents in about a dozen labs and museums.
Skulls and skeletons are only known for Indohyus; for the other genera,
such as Khirtharia and Kunmunella, mostly teeth are known, and it is
possible that some really belong in different artiodactyl families.
The thickened lip of the tympanic bone, the involucrum, gave us a
clue that Indohyus is closely related to whales, but that idea has to be
more formally investigated. Our cladistic analysis (see chapter 10)
shows that whales were indeed more closely related to Indohyus than to
any other artiodactyl, including hippos. Later it was shown that, in
addition, hippos are the closest relatives of the raoellid-cetacean group
(figure 66).5 That result actually does not conflict with the molecular
Before Whales | 201
figure 65. Life reconstruction of the Eocene artiodactyl Indohyus, the closest extinct
relative of whales. Indohyus is in the family Raoellidae, which lived in South Asia from
forty-six to fifty-two million years ago.
data that show that hippos and whales are the closest relatives, because
the molecular studies could not include fossil animals. In other words,
hippos may still be the closest living relative, it is just that extinct Indo-
hyus is even closer. In addition to the presence of an involucrum, these
groups share a number of dental features, such as the front-to-back
arrangement of the upper incisors in the jaw, and the high triangular
202 | Chapter 14
Artiodactyla
Cetacea
even-toed ungulates
whales, dolphins, porpoises
, giraf,
River dolphins
Beluga
Pygmy
, lamas
, sheep
Pygmy
, deer
, peccaries
Gray
right Right Sperm sperm Beaked
and Oceanic
Franciscana
camelspigs cattle
Hippopotamus
whale Rorquals whale whales whale whales whales Susu Baiji Boto
narwhal dolphins Porpoises
modern
antelope
e
eaditsinat
toporidae
Balaenopterida
al
20 million
P
Neobalaenidae
Eschrichtiidae
Balaenidae
Physeteridae
Ziphiidae
Iniidae
Pon
Phocoenidae
Kogiidae
Lipotidae
Monodontidae
Delphinidae
years ago
Baleen whal
Tylopoda
Suina
Ruminantia
es - Mysticeti
40 million
years ago
Basilosauridae
Toothed whales - Odontoceti
Indohyus
Remington Protocetidae
-ocetidae
Ambulocetidae
Pakicetidae
50 million
years ago
archaeocetes - Eocene whales
figure 66. Relationships of cetaceans to artiodactyls. All groups of modern
artiodactyls and cetaceans are included in this figure. Not shown are a large number of
extinct groups that are not discussed in this book.
crowns of the posterior premolars. Tooth-wear patterns of Indohyus
show the same specializations as cetaceans (figure 50).
With that, it seems that the issue of cetacean relations is finally
resolved. Mesonychians are not related to cetaceans. The fossil evidence
shows that cetaceans are derived from a basal, Eocene artiodactyl, and
the closest modern relative of cetaceans is the hippo.
However, that is not the end of the story. Yet another cladistics analy-
sis,6 on a slightly different data-set, confirmed most of the results just
discussed, but found that the support for that view is only slightly stronger
than that for the old mesonychian idea. It is as if those fierce predators are
still waiting in the wings to reclaim their position next to the majestic
whales by pouncing on little vegetarian Indohyus. As one well-known
mammalogist put it: systematics is the soap opera of biology.
With all the rearrangements of the relationships of whales, we have
to wonder whether it would be useful to actually include Indohyus
Before Whales | 203
within Cetacea instead of keeping it just outside the group. After all,
there is nothing sacred about having Pakicetus be the basal cetacean.
And the main feature that characterizes cetaceans, the involucrum, also
occurs in Indohyus. Furthermore, if names should be used for groups
that include an ancestor and all its descendants (monophyletic groups),
the term artiodactyl should now include all cetaceans as well, since they
too descended from the ancestral artiodactyl.
Some authors have indeed advocated for changing the meanings of
Cetacea and Artiodactyla in one way or another,7 but I do not agree
with them. The term artiodactyl has been around for more than 150
years and has had a stable and biologically coherent meaning. Changing
it now would only cause confusion, especially if not all authors follow
the same meaning. It would be especially bad for new students, who
would quickly get confused by names that differ in meaning depending
on when and by whom they are used. My preference is to stick with the
old meaning of the word, where Artiodactyla does not include Cetacea,
and accept the fact that the former does not include all descendants of
that first ancestor. Scientists call this a paraphyletic group, and Artio-
dactyla would be one of those.
Similarly, Cetacea for decades has meant Pakicetus and all its descend-
ants. It too is a biologically coherent group of aquatic predators, albeit
that some had legs and walked. Adding Indohyus to this group muddies
the water: it is so different biologically that it would ren
der the term
Cetacea meaningless in any sense except systematically.
Feeding and Diet. In general, Indohyus has very typical artiodactyl
teeth: a dental formula of 3.1.4.3/3.1.4.3, with upper molars that bear
four cusps, and lower molars that have a high trigonid, with two cusps,
and a low talonid, also with two cusps (figure 34). The shape of these
cusps differs among raoellids: in Indohyus and Kunmunella, the cusps
are sharp and connected by a weak crest, whereas in Khirtharia, the
cusps are low and blunt. Among modern mammals, the first molar type
is common in leaf eaters, while the latter is common in fruit eaters, but
it is not clear whether this difference holds for raoellids. Without doubt,
those dental differences relate to diet and food processing somehow, but
it is unclear how. Stable carbon isotope data indicate that Indohyus and
Khirtharia both fed on land plants.8
Another clue to food processing comes from the relative position of
the joint between lower jaw and skull, where the skull has a socket in
which a ball-like joint, the mandibular condyle, fits. As shown in figure
204 | Chapter 14
25, the condyle is well above the level of the tooth row in herbivores
such as deer. In meat eaters, such as the whales shown in this figure, the
condyle is at the same height as the tooth row. Indohyus’s jaw has an
herbivore’s shape, as expected.
While the molars of Indohyus are not very specialized, their tooth
wear is. Early whales are characterized by nearly exclusive phase I wear
on their lower molars (figure 50, see chapter 11). Eocene artiodactyls
show a combination of phase I, phase II, and apical wear. In Indohyus,
all three wear types are present, but phase I dominates. Apparently, the
land plants that Indohyus ate were processed in ways different from the
way other Eocene artiodactyls processed their food. The dentition also
provides other clues to feeding. Indohyus had a long and pointed snout,
with incisors arranged from front to back, not side to side. This may
have been a specialized mechanism for cropping certain plants. In addi-
tion, its premolars have high crowns with sharp cutting edges on their
sides. At present, the function of these features is not understood, but
with the hundreds of fossils of Indohyus that are known from Kalakot,
and a few more years of study, I am very hopeful that we will know.
Vision and Hearing. The eyes of Indohyus are located on the side of the
skull, as is common in land mammals, and unlike just about all fossil
whales (figure 52). This part of the skull is highly variable and highly
specialized in all Eocene whales, and Indohyus lacks these specializa-
tions. The distance between Indohyus’s orbits and its brain, the inter-
temporal area, is similar to other artiodactyls, and unlike Eocene whales.
The forces related to the continental collision between Asia and India
millions of years ago actually affect how well we can study Indohyus in
the present. Mountain building deformed the rocks and their fossils,
flattening skulls and breaking bones. The skulls of the animal were
crushed, and delicate structures were obliterated. Except for the pres-
ence of an involucrum, very little is known about its ear.
Walking and Swimming. Overall, the skeleton of Indohyus resembles
that of unspecialized artiodactyls, adapted for land locomotion, and
with some specializations often found in runners.9 There are five fingers
and four or five toes, and Indohyus was digitigrade, like a dog, not like
its artiodactyl relatives, who walk on the tips of their toes using hooves
(unguligrade).
In spite of this, there are two lines of evidence that indicate that Indo-
hyus was not a fully terrestrial species. First, some of the bones of Indo-
Before Whales | 205
hyus have a thick outside layer, the cortex, suggesting that one of their
functions is to be ballast while the animal is in the water (figure 62). That
resembles pakicetids, who show this tendency to osteosclerosis to a greater
degree. Also, the oxygen isotopes are interesting. In chapter 9, isotopes
were used to investigate the source of drinking water for some of the early
whales, but here they can help us with another problem. The ratio of 18O
and 16O in the water inside the body of an animal is reflected in its bones
and teeth. Animals lose body water in a number of different ways, for
instance when they pee, and for females, when they produce milk. They
also lose body water when it evaporates through the skin. Interestingly,
evaporation through the skin is a process in which the isotopes are frac-
tionated: water with the lighter oxygen isotope is more likely to go into
the gas phase and disappear from the body than water with the heavier
isotope. As a result, animals that lose a lot of body water through their
skin have an isotopic signature that is skewed toward the heavier isotope.
Mammals that live in water do not sweat or evaporate water, so isotope
ratios can help to discern whether they are aquatic. Indeed, oxygen-iso-
tope values for Indohyus indicate that it spent time in water.
Habitat and Ecology. Indohyus presents a paradox. On the one hand,
carbon-isotope values and the molar shape suggest life on land; on the
other hand, oxygen isotopes and osteosclerosis suggest freshwater. A
possible resolution for the paradox comes from studying a modern
mammal: the mouse deer. Mouse deer live on land, eating the flowers,
leaves, and fruits of terrestrial plants. However, they are always found
near rivers, and when in danger, mouse deer jump into the water and
hide.10 Mouse deer bones are not osteosclerotic, and Indohyus is not
closely related to them. However, they may be the perfect ecological
equivalent. Here, then, may be the key to the origin of aquatic life for
whales: predator-avoidance behavior in their early artiodactyl ancestors.
Kalakot, the fossil site where Indohyus is abundant, has not been
studied sedimentologically, and not much is known about the habitat
these animals were living in. What is known is that there must have been
hundreds of skeletons of Indohyus all washed together buried and
mixed with just a few other forms. Some of the bones found here are
articulated, but most are not. Apparently, there was time for rotting to
disarticulate many of the skeletons. It is possible that this was a flood-
plain of a river with animals living and dying. Skeleto
ns accumulated on
the plain, were dispersed by scavengers, and during the next flood,
washed together into streams.
206 | Chapter 14
a trust for fossils
The pile of rocks holding Indohyus still sits on the estate on Rajpur
Road, with the grave of Friedlinde Obergfell nearby, guarded by Baha-
dur and his wife. The extracted fossils are now safe in the unfinished
house, and we have started to sort through the bags of fossils in the cel-
lar. More fossils are being extracted every day, but the work is slow, and
there is no money to hire a fossil preparator. The trust that manages the
Indohyus fossils has a home, and fossils, and a mission to study them,
but there are no funds to get the study off the ground and save the Indo-
hyus fossils for the future. I hope to avoid it, but it is possible that the
entire place will fold. Sadly, that may be a fitting end to the tragic tale of
Friedlinde Obergfell and Anne Ranga Rao.
Chapter 15
The Way Forward
the big question
I love to talk about whale evolution, and my audiences range from fifth
graders, to our local Rotary club, to cetologists at international meet-
ings. To point out how dramatic the evolution of whales is, I usually
start by asking people to think about two fancy vehicles. I could use a
bullet train and a nuclear submarine, but, because it is less intimidating,
I ask them to think about the Batmobile and the Beatles’ Yellow Subma-
rine. Whales started out with a very elaborately perfected body adapted
to life on land. They changed it, in about eight million years, to a body
perfectly tuned to the ocean. I ask the audience to imagine getting a
team of engineers together to take the Batmobile apart and build the
Yellow Submarine from its parts. Just about everything that works well
on land will fail miserably in water. All the organ systems have to
change—from locomotion, to sense organs, to osmoregulation, to
reproduction. And of course, in evolution, all the intermediate species
were functional in their environment. Adding that requirement would
The Walking Whales Page 31
